System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting existing content of influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each of the influencers using the generative AI model; means for receiving a query from a user; means for selecting an optimal personal AI based on the query; means for transmitting an answer generated by the personal AI to the user; means for collecting feedback from the user; and means for summarizing the feedback and providing the feedback to the influencers.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] The goal is to reduce the burden on influencers, who have to put in a lot of effort to gather information and create content, and to solve the problem of users being unable to immediately access the information they want. Another challenge is to realize an efficient and useful exchange of information between influencers and users in order to effectively utilize user feedback and continuously improve content. [Means for solving the problem]

[0005] This invention provides a means for collecting existing content from influencers and training a generative AI model. This allows for the creation of a personal AI for each influencer, which can receive and process inquiries from users. When a specific influencer is specified, a mechanism is established to select that influencer's personal AI and provide necessary information across the personal AIs of multiple influencers. The system also includes a means for collecting feedback from users, summarizing and analyzing it using the generative AI, and providing it to the influencer, thereby supporting the creation of new content based on user needs.

[0006] An "influencer" is an individual or organization that influences a large number of followers or viewers through their content.

[0007] "Existing content" refers to information resources such as text, images, video, and audio created by influencers that are already publicly available on the Internet or other media.

[0008] A "generative AI model" refers to an artificial intelligence model that uses machine learning techniques to generate natural language like a human for a specific task.

[0009] "Personal AI" refers to artificial intelligence that learns the style and content of a specific influencer and generates responses tailored to that influencer.

[0010] "User" refers to the end user who utilizes an influencer's personal AI to obtain information or provide feedback.

[0011] "Feedback" refers to any questions, comments, reviews, impressions, or other responses or opinions provided by Users.

[0012] "Summarizing" refers to the process of summarizing the feedback in a short and concise manner. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0014] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0015] First, the terms used in the following description will be explained.

[0016] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0017] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0018] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0019] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0020] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0024] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0025] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0026] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0028] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0031] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0032] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0033] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0034] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[0035] Server-side implementation

[0036] 1. Content collection and learning

[0037] The server collects influencers' existing content through APIs and web scraping and stores it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is preprocessed and fed into a generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[0038] 2. Providing chat-style responses

[0039] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[0040] 3. Gathering and Summarizing Feedback

[0041] The server collects feedback from users, including their question history, comments, and impressions. Generative AI is used to summarize the feedback and provide it to influencers in a concise format, allowing the influencers to create new content based on user needs.

[0042] Terminal side embodiment

[0043] 1. Use of the Application

[0044] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[0045] 2. Operations on the chat screen

[0046] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0047] 3. Providing Feedback

[0048] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[0049] User's embodiment

[0050] 1. Start a chat

[0051] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0052] 2. Receiving and using information

[0053] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0054] 3. Providing Feedback

[0055] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using generative AI, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[0056] This will streamline communication between influencers and users and promote the circulation of sustainable content.

[0057] The processing flow will be explained below.

[0058] Server-side processing

[0059] Step 1:

[0060] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[0061] Step 2:

[0062] The server pre-processes the collected content and feeds it into the generative AI model. Pre-processing includes text cleansing (noise removal), tokenization (breaking down words and sentences), and data batching.

[0063] Step 3:

[0064] The server trains a generative AI model to generate a personal AI that reflects the style and expertise of a specific influencer by running feedforward and backpropagation on the model using a large dataset.

[0065] Step 4:

[0066] The server receives inquiries from users via REST APIs, etc. The inquiry content is sent to the server in text format.

[0067] Step 5:

[0068] The server selects an appropriate personal AI based on the user's inquiry. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[0069] Step 6:

[0070] The selected personal AI generates a response to the query, and the generated text is sent to the server and prepared as the final answer.

[0071] Step 7:

[0072] The server then sends the generated answer to the user's device, where it is displayed in the user's application.

[0073] Step 8:

[0074] The server collects feedback from users, which is received in text format.

[0075] Step 9:

[0076] The server uses generative AI to summarize the collected feedback, which involves briefly summarizing the feedback and categorizing positive comments and requests for improvement.

[0077] Step 10:

[0078] The server provides summarized feedback to the influencer, who then uses it to plan and create new content.

[0079] Terminal side processing

[0080] Step 1:

[0081] A system-specific application is installed on the user's device.

[0082] Step 2:

[0083] The user launches the application and logs in by entering their account information.

[0084] Step 3:

[0085] Once the application has logged in, it will display the chat screen.

[0086] Step 4:

[0087] Questions and requests entered by the user are sent from the application to the server.

[0088] Step 5:

[0089] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[0090] Step 6:

[0091] If the user provides feedback, the application sends the feedback to the server.

[0092] User Action

[0093] Step 1:

[0094] The user launches the application and logs in.

[0095] Step 2:

[0096] Users go to the chat screen and type in their questions or requests.

[0097] Step 3:

[0098] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[0099] Step 4:

[0100] When users provide feedback, they type their thoughts about the answer to their question or any follow-up questions into the chat screen.

[0101] Step 5:

[0102] Send feedback and have it delivered to the server.

[0103] This will enable efficient communication between influencers and users, realizing a sustainable circulation of content.

[0104] Example 1

[0105] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0106] Currently, there is a lack of systems that can efficiently collect influencer content and provide personalized information to users. There is also a lack of efficient methods for conveying user feedback to influencers. This makes communication between users and influencers difficult, making it difficult to create new content based on user needs.

[0107] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0108] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiry; means for sending responses generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for storing the collected content in a database; means for using natural language processing technology to preprocess the collected data; means for inputting the preprocessed data as training data into the generative AI model; means for executing the generative AI model to generate responses; and means for summarizing the collected feedback using natural language processing technology. This not only effectively collects existing content from influencers, generates a personal AI based on the collected content, and provides personalized responses to users, but also efficiently communicates user feedback to influencers, facilitating the creation of new content.

[0109] An "influencer" is a person who has influence mainly on social media and online platforms and disseminates their opinions and information to a large number of followers.

[0110] "Content" is a general term for information, including data formats such as text, images, and videos, that are produced and disseminated by influencers.

[0111] A "generative AI model" is an artificial intelligence technology that uses machine learning algorithms to learn from large amounts of data and generate text that is as natural as a human's.

[0112] "Personal AI" is a generative AI model that learns the style and expertise of a specific influencer and generates information similar to that influencer.

[0113] A "query" is a question or request that a user enters into a system to ask for specific information.

[0114] "Feedback" refers to the user's reaction to the information provided, such as rating, comments, or additional questions.

[0115] A "database" is a system for efficiently storing, managing, and searching collected data.

[0116] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, including text cleaning, tokenization, and labeling.

[0117] "Training data" is a pre-processed data set used to train a generative AI model.

[0118] A "REST API" is a type of web service that is an interface for exchanging information between a server and a client using the HTTP protocol.

[0119] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[0120] Server-side implementation

[0121] The server uses APIs and web scraping techniques to collect existing content from influencers. Specifically, it uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The collected data is then stored in a NoSQL database such as MongoDB.

[0122] The server then preprocesses the collected data using natural language processing techniques, such as NLP libraries like SpaCy and NLTK. Preprocessing involves cleaning, tokenizing, and labeling the text. The video data is then converted to text using speech recognition techniques (e.g., Google Cloud Speech-to-Text API).

[0123] The preprocessed data is then used to train a generative AI model, such as GPT-3 or BERT, using machine learning frameworks like PyTorch or TensorFlow to generate a personalized AI model that reflects the style and expertise of a specific influencer.

[0124] The server receives user queries via a REST API. Frameworks such as Flask and FastAPI are used for this. If a user asks, "Tell me about summer fashion coordination," the request is sent to the personal AI, which generates a response. The response is sent back to the server, which then sends it to the user's device.

[0125] In addition, the server collects feedback from users, which is summarized using NLP technology and stored in a database. This summarized feedback is provided to influencers to help them create new content based on user needs.

[0126] Terminal side embodiment

[0127] A mobile application dedicated to this system is installed on the user's device. This application was developed using React Native and Flutter. The user launches the app and logs in. Once authentication is successful, a chat screen appears, allowing the user to freely enter questions or requests.

[0128] When a user types a question into the chat screen, the data is sent from the device to the server. For example, a user might type, "What are the latest summer fashion trends?" and press the send button. This request is sent to the server, which then forwards it to the appropriate personal AI. The generated response is sent from the server to the device and displayed on the chat screen. For example, specific advice such as, "I recommend a linen shirt and a denim skirt" is displayed.

[0129] Users can input their thoughts about the information provided and any follow-up questions as feedback, which is then sent from the device to the server. This feedback also includes a satisfaction rating. The device then sends this information in an appropriate format, making it easier for the server to analyze.

[0130] User's embodiment

[0131] Users interact with the app and initiate a chat to request information from a specific influencer or to find information related to a specific topic, for example, by typing in a question like "summer fashion" and waiting for a response to their question.

[0132] The system receives the answers sent by the influencer AI from the server and utilizes that information. For example, you can receive specific fashion advice such as "Linen shirts and denim skirts are trending."

[0133] Finally, users can provide feedback to the system, including their thoughts and ratings on the information provided. This feedback is then used by the system and influencers to create more optimized content.

[0134] Prompt Sentence Examples

[0135] As a concrete example, suppose a user enters the following question into an application:

[0136] "Please tell me about your summer fashion coordination."

[0137] An example prompt sentence to generate a response to this request:

[0138] "Provide summer fashion recommendations. Users want casual, comfortable outfits."

[0139] By inputting this prompt into a generative AI model, the influencer AI can provide specific suggestions to users, such as "I recommend a linen shirt, a denim skirt, and lightweight sandals."

[0140] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0141] System program processing flow and specific operations

[0142] Step 1:

[0143] The server collects existing content from influencers. It uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The input is the website URL and YouTube video ID, and the output is the extracted text data and downloaded video data. The collected data is stored in MongoDB.

[0144] Step 2:

[0145] The server preprocesses the collected data. It uses an NLP library (e.g., SpaCy or NLTK) to clean, tokenize, and label the text. It then uses the Google Cloud Speech-to-Text API to extract audio from the video data and convert it to text. The input is the raw data collected in step 1, and the output is the preprocessed text data.

[0146] Step 3:

[0147] The server inputs the preprocessed data into a generative AI model and trains the model. For example, it uses PyTorch or TensorFlow to train a model like GPT-3 or BERT. The input is preprocessed text data, and the output is a personal AI model that reflects the style and expertise of a specific influencer.

[0148] Step 4:

[0149] A user launches the application on their device and logs in. The app is developed using React Native and Flutter. If user authentication is successful, a chat screen is displayed. The input is the user's authentication information, and the output is the success / failure status of the authentication and the display of the chat screen.

[0150] Step 5:

[0151] The user enters a question or request on the chat screen, and the device sends the data to the server. For example, a question might be, "Tell me about summer fashion coordination." The input is the text of the question entered by the user, and the output is the request sent to the server.

[0152] Step 6:

[0153] The server sends the received user inquiry to the appropriate personal AI and generates a response. The generated response is received again by the server and sent to the user's device. For example, the personal AI generates an answer such as "I recommend a linen shirt and a denim skirt." The input is the text of the user's inquiry, and the output is the text of the response generated by the personal AI.

[0154] Step 7:

[0155] The terminal receives the response sent from the server and displays it on the chat screen. The user can refer to the displayed information. For example, the advice "We recommend a linen shirt and a denim skirt" is displayed. The input is the text of the response received from the server, and the output is to display that text on the chat screen.

[0156] Step 8:

[0157] The user inputs feedback on the provided information and sends it to the server from the terminal. The feedback includes ratings and follow-up questions. The input is the text of the feedback entered by the user, and the output is the feedback sent to the server.

[0158] Step 9:

[0159] The server summarizes the collected feedback using natural language processing technology and stores it in a database. The summarized feedback is provided to influencers. The input is the text of the feedback received from users, and the output is the summarized text of the feedback. This allows influencers to use it to create new content.

[0160] This will facilitate smooth communication between users and influencers, making it possible to create new content based on user needs.

[0161] (Application example 1)

[0162] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0163] In recent years, there has been an increasing demand for marketing and content provision using influencers. However, there are limited ways for users to receive fast, personalized information from specific influencers, making it difficult for influencers to accurately understand user needs and create new content. Furthermore, existing content provided in the form of videos, images, and text makes it difficult for users to obtain information in real time and lacks interactivity. A new system is needed to solve this issue and streamline communication between users and influencers.

[0164] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0165] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for conducting real-time dialogue via a smartphone or head-mounted display; and means for generating answers in response to questions from users and displaying them in AR or chat format. This allows users to receive information in real time from specific influencer AIs, enabling influencers to create new content based on user needs.

[0166] An "influencer" is an individual or group who influences a large number of followers through an online platform.

[0167] "Existing content" refers to digital information such as text, images, and videos that have already been created and published by influencers.

[0168] A "generative AI model" is an artificial intelligence that is trained using natural language processing and machine learning techniques based on collected data.

[0169] "Personal AI" is a generative AI model that is customized to reflect the style and expertise of a specific influencer.

[0170] An "inquiry" is a question or request for information that a user makes to the influencer AI.

[0171] A "smartphone" is a portable electronic device that, in addition to the functions of a mobile phone, has advanced computing power and Internet access capabilities.

[0172] A "head-mounted display" is a display device worn on the user's head, typically for displaying visual and auditory content.

[0173] "Real-time" refers to processing and responding to information and data instantly, without delay.

[0174] "Chat format" refers to a format in which a conversation is carried out between the user and the system by sending and receiving text messages.

[0175] "AR (Augmented Reality)" is a technology that displays digital information overlaid on images of the real world.

[0176] "Feedback" refers to reaction information based on comments, impressions, ratings, etc. provided by users.

[0177] The present invention is a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in chat or AR formats. The following describes an embodiment of the system in detail.

[0178] Server-side implementation

[0179] 1. Content collection and learning

[0180] The server uses APIs and web scraping technology to collect influencers' existing content. This content includes video data from video platforms and text and image data from social media. This data is stored in a database and, after preprocessing, is used as training data for the generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[0181] 2. Providing chat and AR responses

[0182] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer based on the inquiry, and the server sends the answer to the user's device. The answer is displayed on the user's device in chat format or AR format.

[0183] 3. Gathering and Summarizing Feedback

[0184] The server collects feedback from users, including their question history, comments, and impressions. It then uses a generative AI model to summarize the feedback and provide it to influencers, who can then create new content based on user needs.

[0185] Terminal side embodiment

[0186] 1. Use of the Application

[0187] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. If a head-mounted display is used, information is provided in AR format.

[0188] 2. Operations on the chat screen and AR screen

[0189] The questions entered by the user are sent from the device to the server. The responses received from the server are instantly displayed on the chat screen or AR screen, and the user can use the information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0190] 3. Providing Feedback

[0191] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[0192] User's embodiment

[0193] 1. Start a chat

[0194] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0195] 2. Receiving and using information

[0196] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0197] 3. Providing Feedback

[0198] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using a generative AI model, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[0199] Specific examples

[0200] If the user asks "Tell me about summer fashion," the following prompt will be generated:

[0201] "Users ask AI fashion influencers questions: Tell us about summer fashion"

[0202] This allows users to receive specific fashion advice in chat or AR format.

[0203] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0204] Step 1:

[0205] The server collects influencers' existing content. Specifically, it uses APIs and web scraping technology to collect text, image, and video data from video platforms and social media, and stores it in a database. The collected data undergoes preprocessing (noise removal, normalization, tokenization, etc.) and is prepared as training data for the generative AI model. The input is public content on the internet, and the output is preprocessed learning data.

[0206] Step 2:

[0207] The server uses the preprocessed data to train a generative AI model using natural language processing or machine learning algorithms (e.g., the GPT series). The input is the preprocessed content data, and the output is a generative AI model that has learned the style and expertise of a specific influencer. Specifically, the server splits the dataset and iteratively trains the model.

[0208] Step 3:

[0209] The server receives inquiries from users via a REST API. For example, if a user asks, "Tell me about summer fashion coordination," the server converts the inquiry into an appropriate format. The input is the user's inquiry text, and the output is data that has been appropriately analyzed from the inquiry. Specifically, the server uses natural language processing to tokenize and analyze the inquiry.

[0210] Step 4:

[0211] The server selects the optimal personal AI based on the content of the inquiry. This generates a response based on the style and expertise of a specific influencer. The input is the analyzed inquiry content and information from multiple personal AIs, and the output is the optimal personal AI. Specifically, it compares the characteristics of the inquiry content and the personal AIs and selects the AI ​​with the highest degree of match.

[0212] Step 5:

[0213] The server generates a response to the inquiry using the selected personal AI. The generated answer is sent to the user. The input is the inquiry content and the selected personal AI, and the output is the generated answer. Specifically, a prompt sentence is input into the generative AI model to obtain a natural conversational response.

[0214] Step 6:

[0215] The device displays the answers received from the server on the chat screen or in AR format. The input is the answer data sent from the server, and the output is the information visually displayed to the user. Specifically, the information is displayed in text format or augmented reality format on the screen of the chat application or AR display.

[0216] Step 7:

[0217] The user inputs feedback on the provided information. The input is the user's feedback text, and the output is the feedback data sent to the server. Specifically, the information is entered and sent from a feedback input form in the chat application.

[0218] Step 8:

[0219] The server collects feedback from users and summarizes it using a generative AI model. The input is the feedback data sent by the user, and the output is the summarized feedback information. Specifically, the server analyzes the feedback data using natural language processing technology, extracts key points, and generates a summary.

[0220] Step 9:

[0221] The server provides summarized feedback to influencers. The input is the summarized feedback data, and the output is the data provided in the form of notifications or reports to the influencers. Specific actions include sending the feedback to the influencers' dashboards or notification systems.

[0222] The above are the specific processing steps and their contents of the system that realizes the application example.

[0223] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0224] The present invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, provides information to users in a chat format, and combines it with an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions. The following describes an embodiment of the system in detail.

[0225] Server-side implementation

[0226] 1. Content collection and learning

[0227] The server first collects the influencer's existing content through APIs and web scraping, storing it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is then preprocessed and fed into a generative AI model. The generative AI model then learns from this data and generates a personal AI that reflects the style and expertise of the specific influencer.

[0228] 2. Providing chat-style responses

[0229] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[0230] 3. Emotion engine integration

[0231] The server is equipped with an emotion engine that analyzes user inquiries and feedback to recognize user emotions, such as joy, anger, sadness, and surprise, based on the context of the text.

[0232] 4. Emotion-Based Response Modulation

[0233] The server adjusts the personal AI's response based on the user's emotional information analyzed by the emotion engine. For example, if a user gives negative feedback such as "This idea was completely useless," the emotion engine will determine this as "anger," and the personal AI will adjust its response to be more polite, such as providing additional advice or citing past success stories.

[0234] 5. Collecting and Summarizing Feedback

[0235] The server collects user feedback and uses generative AI to summarize it. The summary, including the sentiment data of the feedback, is provided to influencers in a concise format, allowing them to create new content based on user needs.

[0236] Terminal side embodiment

[0237] 1. Use of the Application

[0238] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[0239] 2. Operations on the chat screen

[0240] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0241] 3. Providing Feedback

[0242] The feedback provided by the user is transmitted from the device to the server. This feedback includes new questions, comments, satisfaction ratings, and even emotional data. The device transmits this information in an appropriate format to facilitate analysis on the server side.

[0243] User's embodiment

[0244] 1. Start a chat

[0245] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0246] 2. Receiving and using information

[0247] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0248] 3. Receiving emotional responses

[0249] Based on the user's input query and feedback, the system uses an emotion engine to generate an appropriate response. For example, if the user is confused, the system will provide additional explanation.

[0250] 4. Providing Feedback

[0251] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server then collects this feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows influencers to more accurately understand users' needs and emotions and improve their content.

[0252] This will streamline communication between influencers and users, and promote the circulation of more personalized and sustainable content by taking emotions into consideration.

[0253] The processing flow will be explained below.

[0254] Server-side processing

[0255] Step 1:

[0256] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[0257] Step 2:

[0258] The server pre-processes the collected content, specifically cleansing, tokenizing, and batching the text data to prepare it for ingestion into the generative AI model.

[0259] Step 3:

[0260] The server trains a generative AI model, using the pre-processed content as training data and running feedforward and backpropagation on the model to generate a personal AI that reflects the style and expertise of a particular influencer.

[0261] Step 4:

[0262] The server receives inquiries from users via the REST API. For example, the user might ask, "Tell me about summer fashion coordination."

[0263] Step 5:

[0264] The server uses an emotion engine to analyze the content of the user's inquiry and recognize their emotion. For example, it can determine their interest from the context of "Please tell me."

[0265] Step 6:

[0266] The server selects an appropriate personal AI based on the user's inquiry and emotional information. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[0267] Step 7:

[0268] The selected personal AI generates an answer based on the user's query and emotional information, and the generated text is sent to the server to be prepared as the final answer.

[0269] Step 8:

[0270] The server then sends the generated answer to the user's device, such as "A cool linen shirt and denim skirt are recommended for this summer."

[0271] Step 9:

[0272] The server collects feedback from users. The feedback is received in text format, and the emotion engine simultaneously analyzes the emotion data. For example, feedback such as "This was helpful" is recognized as "satisfied / positive."

[0273] Step 10:

[0274] The server summarizes the collected feedback and its sentiment data, then uses generative AI to summarize it and provide it to influencers in a concise format, allowing them to create new content that reflects user needs and sentiment.

[0275] Terminal side processing

[0276] Step 1:

[0277] The terminal installs a system-specific application.

[0278] Step 2:

[0279] The user launches the application and logs in by entering their account information.

[0280] Step 3:

[0281] Once the application has logged in, it will display the chat screen.

[0282] Step 4:

[0283] Questions and requests entered by the user are sent from the device to the server.

[0284] Step 5:

[0285] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[0286] Step 6:

[0287] When the user provides feedback, the application sends the feedback to the server, along with the user's emotional data.

[0288] User Action

[0289] Step 1:

[0290] The user launches the application and logs in.

[0291] Step 2:

[0292] Users go to the chat screen and type in their questions or requests.

[0293] Step 3:

[0294] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[0295] Step 4:

[0296] When users provide feedback, they type their thoughts about the answer or follow-up questions into the chat window, and emotional data is automatically analyzed and sent along with the feedback.

[0297] Step 5:

[0298] Send feedback and have it delivered to the server, where it will be analyzed and summarized by the system and provided to the influencer.

[0299] This will enable more efficient communication between influencers and users, and by taking emotions into consideration, will enable a more personalized and sustainable circulation of content.

[0300] Example 2

[0301] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0302] In the conventional system, information provided through influencer content was one-way, making it difficult to respond flexibly to user emotions. Furthermore, there was no system in place to efficiently collect and analyze user feedback, making it difficult for influencers to accurately understand users' needs and emotions. This limited the user experience and prevented effective communication between influencers and users.

[0303] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0304] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the users; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for preprocessing text data; means for analyzing emotions that recognize the user's emotions; and means for adjusting response content based on the recognized emotions. This enables flexible information provision based on user emotions and efficient collection and analysis of feedback.

[0305] An "influencer" is an individual or organization that uses social media or other digital platforms to influence a large number of followers.

[0306] "Existing content" refers to digital media data such as videos, text, and images that have already been published by influencers.

[0307] "Collection Methods" refers to the technical means used to obtain existing content from influencers, using technologies such as APIs and web scraping.

[0308] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of data and generates natural language and images.

[0309] "Training method" refers to the technical process or method for training a generative AI model based on existing content.

[0310] "Personal AI" refers to artificial intelligence that learns the style and expertise of a specific influencer and provides information tailored to that influencer.

[0311] "Means for receiving inquiries" refers to the technical means by which the server receives inquiries from users, such as through a REST API.

[0312] "Means for selecting the optimal personal AI" refers to algorithms and technologies for selecting the appropriate personal AI based on the content of the user's inquiry.

[0313] "Means for sending answers" refers to the communication means for sending answers generated by personal AI to the user's device.

[0314] "Feedback Collection Means" refers to the technical means by which the Server collects comments and ratings provided by Users.

[0315] "Means for summarizing and providing feedback" refers to technology for summarizing collected feedback using a generative AI model and providing that summary to influencers.

[0316] "Means for preprocessing text data" refers to the technical means for analyzing and cleansing collected text data and converting it into a form that is easy for generative AI models to use.

[0317] "Sentiment analysis methods" refer to algorithms and technologies used to recognize and classify emotions from user text data.

[0318] "Means for adjusting response content" refers to technical means for appropriately changing the responses of personal AI based on the results of emotion analysis.

[0319] This invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, and provides information to users in chat format, as well as a system that combines an emotion engine that recognizes the user's emotions and adjusts responses based on them.

[0320] Server-side implementation

[0321] 1. Content collection and learning

[0322] The server first collects the influencer's existing content through APIs and web scraping. For example, it obtains video data from video platforms and collects text and image data from social networking services. This data is preprocessed and fed into the generative AI model. Preprocessing includes converting the video's audio to text, extracting features from the image data, and cleansing the text. The generative AI model then learns from this data to generate a personal AI that reflects the style and expertise of a specific influencer. For example, it obtains video data using YouTube's API, extracts audio from the video using FFmpeg, and converts it into text using the Google Cloud Speech-to-Text API. The collected text data is then formatted using natural language processing technology.

[0323] 2. Receiving and Processing User Requests

[0324] When a user makes a request through the application, such as "Tell me about summer fashion coordination," the request is sent to the server via a REST API. The server selects an appropriate personal AI and communicates the inquiry to that personal AI. The generative AI model generates an answer in response to the inquiry. If the answer is "I recommend a linen shirt and a denim skirt," the server sends this answer to the user's device.

[0325] 3. Emotion engine integration and emotion analysis

[0326] The server is integrated with an emotion engine that analyzes user inquiries and feedback to recognize emotions. For example, if a user sends feedback such as "This idea was completely useless," the emotion engine analyzes the text and determines that the emotion is "anger." This is done using natural language processing technology and emotion classification algorithms.

[0327] 4. Adjusting the response

[0328] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We apologize for the inconvenience. What do you think of these suggestions?" and sends it to the user.

[0329] 5. Collecting and Summarizing Feedback

[0330] The server collects user feedback and uses generative AI to summarize it, including emotional information, and provides it to influencers, who can then create new content based on user needs.

[0331] Terminal side embodiment

[0332] 1. Use of the Application

[0333] Users install the dedicated application and log in. After logging in, a chat screen appears, allowing users to freely post questions or requests.

[0334] 2. Operations on the chat screen

[0335] When a user enters a question, the content is sent from the device to the server. The server's response is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0336] 3. Providing Feedback

[0337] Users can enter feedback on the information provided and send it from their device to the server. This feedback can include new questions, comments, satisfaction ratings, and sentiment data. The device then sends this information in an appropriate format to facilitate analysis by the server.

[0338] User's embodiment

[0339] 1. Start a chat

[0340] Users can navigate the app to start a chat to request information from a specific influencer or find information related to a specific topic, for example, by typing, "Tell me about summer fashion."

[0341] 2. Receiving and using information

[0342] The system receives the personal AI's response from the server and utilizes that information. For example, it receives specific advice such as, "Recommended fashion items for summer are a linen shirt and a denim skirt."

[0343] 3. Receiving emotional responses

[0344] When a user enters a query or feedback, the emotion engine recognizes the user's emotions and generates an appropriate response. For example, if the user expresses dissatisfaction by saying "this advice was not helpful," the engine will provide additional advice.

[0345] 4. Providing Feedback

[0346] Users can input their thoughts on the information provided or any follow-up questions as feedback and provide it to the server. The server then collects the feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows the influencer to more accurately understand the needs and sentiments of users and improve their content.

[0347] Examples of prompt statements

[0348] "Please tell me about your summer fashion coordination."

[0349] "This idea was completely useless."

[0350] "Please tell me some more casual outfit ideas."

[0351] This will streamline communication between influencers and users, and by taking emotions into consideration, more personalized information will be provided.

[0352] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0353] Step 1:

[0354] Content Collection

[0355] The server collects influencers' existing content through APIs and web scraping. Specifically, it uses the YouTube API to obtain video data and social media APIs to collect text and image data. The server specifies the influencer's account or specific keywords as input, and obtains video data, text data, and image data as output.

[0356] What it does: The server periodically sends a YouTube API request to get the metadata and video URL of new videos, then uses another API request to download the videos.

[0357] Step 2:

[0358] Data Preprocessing

[0359] The server preprocesses the acquired data. Specifically, it converts the video audio into text (e.g., using FFmpeg), extracts features from the image data (e.g., using TensorFlow), and cleanses the text data (e.g., removing unnecessary characters and tags using regular expressions). It receives video, text, and image data as input and generates preprocessed data as output.

[0360] What it does: It converts the video into an audio file using FFmpeg, then converts the audio to text using the Google Cloud Speech-to-Text API, and then formats the text using a natural language processing library.

[0361] Step 3:

[0362] Training generative AI models

[0363] The server uses the preprocessed data to train a generative AI model. Specifically, it uses GPT-3, for example, to learn style and expertise from influencer content. It receives the preprocessed data as input and generates a generative AI model as output.

[0364] Specific operation: The server uses TensorFlow or PyTorch to train a generative AI model and saves the model after learning is complete.

[0365] Step 4:

[0366] Receiving a user request

[0367] The server receives user inquiries via a REST API. For example, it receives a request such as, "Tell me about summer fashion coordination." It receives the user's question as input and sends the request to the appropriate personal AI as output.

[0368] What happens: The user's request is sent to the server as an HTTP POST request and parsed.

[0369] Step 5:

[0370] Generate and send a response

[0371] The server passes the received request to the generative AI model, which then generates an appropriate answer. For example, it might generate an answer such as "I recommend a linen shirt and a denim skirt," and the server sends that answer to the user's device. The generative AI model receives the user request and the generated answer as input and sends the generated answer to the user as output.

[0372] How it works: The generative AI model generates a response, and the server displays the response on the user's chat screen in real time.

[0373] Step 6:

[0374] Emotion analysis

[0375] The server uses an emotion engine to analyze the text of user inquiries and feedback and recognize the user's emotions. For example, if the user gives negative feedback such as "This idea was completely useless," it will identify it as "anger." It receives user feedback as input and generates emotion data as output.

[0376] Specific operation: Extracts emotions from text using a sentiment analysis algorithm and assigns sentiment tags.

[0377] Step 7:

[0378] Regulating responses based on emotions

[0379] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We're sorry for the inconvenience. What do you think of these suggestions?" and sends it to the user. It receives the sentiment data and the original answer as input and generates the adjusted response as output.

[0380] Specific operation: Based on the emotion data, a new prompt is given to the generative AI model to generate an appropriate response.

[0381] Step 8:

[0382] Collecting and summarizing feedback

[0383] The server collects user feedback and uses generative AI to summarize it. The feedback includes emotional information and provides it to influencers, who can then create new content based on user needs. The server receives user feedback as input and generates summarized feedback as output.

[0384] Specific operation: Collect multiple feedback data, generate a summary using natural language processing technology, and notify the influencer of the summary.

[0385] (Application example 2)

[0386] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0387] In recent years, as the influence of influencers has grown, there has been an increasing demand for information provision systems that utilize influencer content. However, existing systems have difficulty generating appropriate responses that take user emotions into account, limiting the improvement of user experience. There is also a growing need for systems that not only generate personal AI that reflects the style of a specific influencer, but also have the ability to adjust responses according to different emotional states. Furthermore, there is a need for a mechanism that allows users to receive more personalized information in virtual stores. The present invention aims to solve these problems and provide a high-quality information provision system that reflects user emotions.

[0388] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting existing content of influencers, means for training a generative AI model based on the collected existing content, means for generating a personal AI for each influencer using the generative AI model, means for receiving inquiries from users, means for selecting an optimal personal AI based on the inquiries, means for sending answers generated by the personal AI to the user, means for collecting feedback from users, means for summarizing the feedback and providing it to the influencer, means for analyzing the user's emotions, and means for adjusting the response of the personal AI based on the analyzed emotions. This makes it possible to adjust responses according to the user's emotional state, thereby enabling the provision of more personalized, high-quality information.

[0389] An "influencer" is someone who has a large following and whose opinions and actions influence other people.

[0390] "Existing content" refers to data such as text, images, and videos that influencers have previously posted.

[0391] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task based on collected data.

[0392] "Personal AI" refers to artificial intelligence generated to reflect the style and expertise of a specific influencer.

[0393] "Enquiry" means a question or request that a user enters into a system seeking information.

[0394] A "response" is an answer that a personal AI provides to a user's inquiry.

[0395] "Feedback" refers to the evaluations and opinions that users give in response to suggestions and advice.

[0396] "Sentiment analysis" is the process of analyzing and identifying a user's emotional state from their text and behavior.

[0397] "Response adjustment" is the process of appropriately modifying the answers provided by personal AI based on analyzed emotional data.

[0398] This invention is a system that collects existing influencer content, creates a generative AI model, and generates a personal AI for each influencer, which provides appropriate responses to user inquiries. It also analyzes user emotions and adjusts responses based on that information, thereby providing more personalized information.

[0399] Server-side implementation

[0400] The server first collects existing content from influencers through APIs and web scraping. The collected data is stored in a database, where it undergoes pre-processing before being fed into a generative AI model. This generative AI model learns the style and expertise of a specific influencer and generates a personal AI based on the results.

[0401] Inquiries from users are received via REST APIs, and the most suitable personal AI is selected based on the content of the inquiry. The selected personal AI generates a response to the inquiry and sends it from the server to the user's device. At this time, the server is equipped with an emotion analysis engine that analyzes the user's emotions and adjusts the response based on that information.

[0402] By utilizing the sentiment analysis engine, if a user sends negative feedback, such as "This advice was not helpful," the sentiment engine will detect "anger" and the personal AI will adjust its response to be more polite. This response adjustment process may include providing additional advice or citing past success stories.

[0403] Terminal side embodiment

[0404] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. Questions entered by the user are sent from the device to the server, and responses received from the server are immediately displayed on the chat screen.

[0405] The feedback sent from the device includes satisfaction ratings and emotional data. This information is sent to a server for analysis. The emotional data is fed back into the generative AI model, helping to improve the accuracy of each influencer's personal AI.

[0406] User's embodiment

[0407] Users operate the application and start a chat when they want information from a specific influencer. They receive a response to their input from a personal AI, which then utilizes that information. For example, if a user asks, "I'd like some advice on casual summer style," they can receive specific advice from the influencer AI, such as, "I recommend a linen shirt and a denim skirt."

[0408] The feedback provided by users includes information based on sentiment analysis. For example, if a user is confused and says, "I want more specific suggestions," the system will provide additional, detailed advice. This feedback is collected on the server and provided to influencers, allowing them to generate new content based on user needs.

[0409] Examples and prompts

[0410] As a concrete example, imagine a scenario in which a user is choosing fashion in a virtual store. For example, if the user asks, "I'd like some advice on a casual summer style," the system will respond, "I recommend a linen shirt and a denim skirt."

[0411] Example prompt sentence:

[0412] "A user is asking about casual summer style. What suggestions can you offer?"

[0413] "The user is confused. Please provide more specific advice."

[0414] This allows them to take into account the user's emotions and provide a more personalized response.

[0415] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0416] Step 1:

[0417] The server collects existing content from influencers through APIs and web scraping, and stores it in a database in the form of video, text, images, etc.

[0418] Input: Influencer content (e.g., video, text, images)

[0419] Output: Content data stored in the database

[0420] How it works: The server uses web scraping tools and APIs to retrieve content from social media and video platforms and stores it in a database.

[0421] Step 2:

[0422] The server preprocesses the collected data and feeds it into the generative AI model. Data preprocessing includes removing unnecessary information and standardizing the format.

[0423] Input: Content data in the database

[0424] Output: Preprocessed data

[0425] Specific operation: The server organizes the data using data cleaning techniques and performs preprocessing such as tokenizing text and resizing images.

[0426] Step 3:

[0427] The server uses the preprocessed data to train a generative AI model, which acts as a personal AI.

[0428] Input: Preprocessed data

[0429] Output: A trained generative AI model

[0430] How it works: The server runs machine learning algorithms and trains generative AI models based on the collected data.

[0431] Step 4:

[0432] The user's device starts the application and displays the login screen. The user logs in.

[0433] Input: User login information

[0434] Output: User authentication result

[0435] Specific operation: The terminal sends the login information entered by the user to the server, and the server performs authentication.

[0436] Step 5:

[0437] The terminal displays a chat screen and receives inquiries from users.

[0438] Input: User's inquiry

[0439] Output: Inquiry content displayed on the chat screen

[0440] Specific behavior: Sends the text entered by the user on the chat screen to the server.

[0441] Step 6:

[0442] The server receives the user's inquiry and selects the most suitable personal AI.

[0443] Input: User's inquiry

[0444] Output: Selected personal AI

[0445] Specific operation: The server analyzes the query content and selects the most suitable personal AI model.

[0446] Step 7:

[0447] The server sends the response generated by the selected personal AI to the user's device.

[0448] Input: Enquiry details, personal AI model

[0449] Output: Response by personal AI

[0450] Specific operation: The personal AI generates a response based on the inquiry, and the server sends that response to the user's device.

[0451] Step 8:

[0452] The server receives the user's feedback and analyzes its content and sentiment.

[0453] Input: User feedback

[0454] Output: Parsed emotion data

[0455] Specific operation: The emotion analysis engine built into the server analyzes the feedback text to determine the emotion.

[0456] Step 9:

[0457] The server adjusts the personal AI's responses based on the analyzed emotions.

[0458] Input: Emotional data, Personal AI response

[0459] Output: Adjusted response

[0460] Specific behavior: The server takes into account the emotional data and appropriately modifies the personal AI's responses, for example, making the response more polite if anger is detected.

[0461] Step 10:

[0462] The server summarizes the user's feedback and stores it for providing to the influencer.

[0463] Input: User feedback

[0464] Output: Summarized feedback data

[0465] Specific operation: The server uses generative AI to summarize the feedback content and saves the results in a database.

[0466] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0467] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0468] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0469] [Second embodiment]

[0470] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0471] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0472] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0473] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0474] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0475] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0476] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0477] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0478] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0479] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0480] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0481] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0482] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[0483] Server-side implementation

[0484] 1. Content collection and learning

[0485] The server collects influencers' existing content through APIs and web scraping and stores it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is preprocessed and fed into a generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[0486] 2. Providing chat-style responses

[0487] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[0488] 3. Gathering and Summarizing Feedback

[0489] The server collects feedback from users, including their question history, comments, and impressions. Generative AI is used to summarize the feedback and provide it to influencers in a concise format, allowing the influencers to create new content based on user needs.

[0490] Terminal side embodiment

[0491] 1. Use of the Application

[0492] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[0493] 2. Operations on the chat screen

[0494] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0495] 3. Providing Feedback

[0496] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[0497] User's embodiment

[0498] 1. Start a chat

[0499] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0500] 2. Receiving and using information

[0501] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0502] 3. Providing Feedback

[0503] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using generative AI, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[0504] This will streamline communication between influencers and users and promote the circulation of sustainable content.

[0505] The processing flow will be explained below.

[0506] Server-side processing

[0507] Step 1:

[0508] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[0509] Step 2:

[0510] The server pre-processes the collected content and feeds it into the generative AI model. Pre-processing includes text cleansing (noise removal), tokenization (breaking down words and sentences), and data batching.

[0511] Step 3:

[0512] The server trains a generative AI model to generate a personal AI that reflects the style and expertise of a specific influencer by running feedforward and backpropagation on the model using a large dataset.

[0513] Step 4:

[0514] The server receives inquiries from users via REST APIs, etc. The inquiry content is sent to the server in text format.

[0515] Step 5:

[0516] The server selects an appropriate personal AI based on the user's inquiry. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[0517] Step 6:

[0518] The selected personal AI generates a response to the query, and the generated text is sent to the server and prepared as the final answer.

[0519] Step 7:

[0520] The server then sends the generated answer to the user's device, where it is displayed in the user's application.

[0521] Step 8:

[0522] The server collects feedback from users, which is received in text format.

[0523] Step 9:

[0524] The server uses generative AI to summarize the collected feedback, which involves briefly summarizing the feedback and categorizing positive comments and requests for improvement.

[0525] Step 10:

[0526] The server provides summarized feedback to the influencer, who then uses it to plan and create new content.

[0527] Terminal side processing

[0528] Step 1:

[0529] A system-specific application is installed on the user's device.

[0530] Step 2:

[0531] The user launches the application and logs in by entering their account information.

[0532] Step 3:

[0533] Once the application has logged in, it will display the chat screen.

[0534] Step 4:

[0535] Questions and requests entered by the user are sent from the application to the server.

[0536] Step 5:

[0537] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[0538] Step 6:

[0539] If the user provides feedback, the application sends the feedback to the server.

[0540] User Action

[0541] Step 1:

[0542] The user launches the application and logs in.

[0543] Step 2:

[0544] Users go to the chat screen and type in their questions or requests.

[0545] Step 3:

[0546] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[0547] Step 4:

[0548] When users provide feedback, they type their thoughts about the answer to their question or any follow-up questions into the chat screen.

[0549] Step 5:

[0550] Send feedback and have it delivered to the server.

[0551] This will enable efficient communication between influencers and users, realizing a sustainable circulation of content.

[0552] Example 1

[0553] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0554] Currently, there is a lack of systems that can efficiently collect influencer content and provide personalized information to users. There is also a lack of efficient methods for conveying user feedback to influencers. This makes communication between users and influencers difficult, making it difficult to create new content based on user needs.

[0555] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0556] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiry; means for sending responses generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for storing the collected content in a database; means for using natural language processing technology to preprocess the collected data; means for inputting the preprocessed data as training data into the generative AI model; means for executing the generative AI model to generate responses; and means for summarizing the collected feedback using natural language processing technology. This not only effectively collects existing content from influencers, generates a personal AI based on the collected content, and provides personalized responses to users, but also efficiently communicates user feedback to influencers, facilitating the creation of new content.

[0557] An "influencer" is a person who has influence mainly on social media and online platforms and disseminates their opinions and information to a large number of followers.

[0558] "Content" is a general term for information, including data formats such as text, images, and videos, that are produced and disseminated by influencers.

[0559] A "generative AI model" is an artificial intelligence technology that uses machine learning algorithms to learn from large amounts of data and generate text that is as natural as a human's.

[0560] "Personal AI" is a generative AI model that learns the style and expertise of a specific influencer and generates information similar to that influencer.

[0561] A "query" is a question or request that a user enters into a system to ask for specific information.

[0562] "Feedback" refers to the user's reaction to the information provided, such as rating, comments, or additional questions.

[0563] A "database" is a system for efficiently storing, managing, and searching collected data.

[0564] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, including text cleaning, tokenization, and labeling.

[0565] "Training data" is a pre-processed data set used to train a generative AI model.

[0566] A "REST API" is a type of web service that is an interface for exchanging information between a server and a client using the HTTP protocol.

[0567] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[0568] Server-side implementation

[0569] The server uses APIs and web scraping techniques to collect existing content from influencers. Specifically, it uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The collected data is then stored in a NoSQL database such as MongoDB.

[0570] The server then preprocesses the collected data using natural language processing techniques, such as NLP libraries like SpaCy and NLTK. Preprocessing involves cleaning, tokenizing, and labeling the text. The video data is then converted to text using speech recognition techniques (e.g., Google Cloud Speech-to-Text API).

[0571] The preprocessed data is then used to train a generative AI model, such as GPT-3 or BERT, using machine learning frameworks like PyTorch or TensorFlow to generate a personalized AI model that reflects the style and expertise of a specific influencer.

[0572] The server receives user queries via a REST API. Frameworks such as Flask and FastAPI are used for this. If a user asks, "Tell me about summer fashion coordination," the request is sent to the personal AI, which generates a response. The response is sent back to the server, which then sends it to the user's device.

[0573] In addition, the server collects feedback from users, which is summarized using NLP technology and stored in a database. This summarized feedback is provided to influencers to help them create new content based on user needs.

[0574] Terminal side embodiment

[0575] A mobile application dedicated to this system is installed on the user's device. This application was developed using React Native and Flutter. The user launches the app and logs in. Once authentication is successful, a chat screen appears, allowing the user to freely enter questions or requests.

[0576] When a user types a question into the chat screen, the data is sent from the device to the server. For example, a user might type, "What are the latest summer fashion trends?" and press the send button. This request is sent to the server, which then forwards it to the appropriate personal AI. The generated response is sent from the server to the device and displayed on the chat screen. For example, specific advice such as, "I recommend a linen shirt and a denim skirt" is displayed.

[0577] Users can input their thoughts about the information provided and any follow-up questions as feedback, which is then sent from the device to the server. This feedback also includes a satisfaction rating. The device then sends this information in an appropriate format, making it easier for the server to analyze.

[0578] User's embodiment

[0579] Users interact with the app and initiate a chat to request information from a specific influencer or to find information related to a specific topic, for example, by typing in a question like "summer fashion" and waiting for a response to their question.

[0580] The system receives the answers sent by the influencer AI from the server and utilizes that information. For example, you can receive specific fashion advice such as "Linen shirts and denim skirts are trending."

[0581] Finally, users can provide feedback to the system, including their thoughts and ratings on the information provided. This feedback is then used by the system and influencers to create more optimized content.

[0582] Prompt Sentence Examples

[0583] As a concrete example, suppose a user enters the following question into an application:

[0584] "Please tell me about your summer fashion coordination."

[0585] An example prompt sentence to generate a response to this request:

[0586] "Provide summer fashion recommendations. Users want casual, comfortable outfits."

[0587] By inputting this prompt into a generative AI model, the influencer AI can provide specific suggestions to users, such as "I recommend a linen shirt, a denim skirt, and lightweight sandals."

[0588] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0589] System program processing flow and specific operations

[0590] Step 1:

[0591] The server collects existing content from influencers. It uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The input is the website URL and YouTube video ID, and the output is the extracted text data and downloaded video data. The collected data is stored in MongoDB.

[0592] Step 2:

[0593] The server preprocesses the collected data. It uses an NLP library (e.g., SpaCy or NLTK) to clean, tokenize, and label the text. It then uses the Google Cloud Speech-to-Text API to extract audio from the video data and convert it to text. The input is the raw data collected in step 1, and the output is the preprocessed text data.

[0594] Step 3:

[0595] The server inputs the preprocessed data into a generative AI model and trains the model. For example, it uses PyTorch or TensorFlow to train a model like GPT-3 or BERT. The input is preprocessed text data, and the output is a personal AI model that reflects the style and expertise of a specific influencer.

[0596] Step 4:

[0597] A user launches the application on their device and logs in. The app is developed using React Native and Flutter. If user authentication is successful, a chat screen is displayed. The input is the user's authentication information, and the output is the success / failure status of the authentication and the display of the chat screen.

[0598] Step 5:

[0599] The user enters a question or request on the chat screen, and the device sends the data to the server. For example, a question might be, "Tell me about summer fashion coordination." The input is the text of the question entered by the user, and the output is the request sent to the server.

[0600] Step 6:

[0601] The server sends the received user inquiry to the appropriate personal AI and generates a response. The generated response is received again by the server and sent to the user's device. For example, the personal AI generates an answer such as "I recommend a linen shirt and a denim skirt." The input is the text of the user's inquiry, and the output is the text of the response generated by the personal AI.

[0602] Step 7:

[0603] The terminal receives the response sent from the server and displays it on the chat screen. The user can refer to the displayed information. For example, the advice "We recommend a linen shirt and a denim skirt" is displayed. The input is the text of the response received from the server, and the output is to display that text on the chat screen.

[0604] Step 8:

[0605] The user inputs feedback on the provided information and sends it to the server from the terminal. The feedback includes ratings and follow-up questions. The input is the text of the feedback entered by the user, and the output is the feedback sent to the server.

[0606] Step 9:

[0607] The server summarizes the collected feedback using natural language processing technology and stores it in a database. The summarized feedback is provided to influencers. The input is the text of the feedback received from users, and the output is the summarized text of the feedback. This allows influencers to use it to create new content.

[0608] This will facilitate smooth communication between users and influencers, making it possible to create new content based on user needs.

[0609] (Application example 1)

[0610] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0611] In recent years, there has been an increasing demand for marketing and content provision using influencers. However, there are limited ways for users to receive fast, personalized information from specific influencers, making it difficult for influencers to accurately understand user needs and create new content. Furthermore, existing content provided in the form of videos, images, and text makes it difficult for users to obtain information in real time and lacks interactivity. A new system is needed to solve this issue and streamline communication between users and influencers.

[0612] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0613] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for conducting real-time dialogue via a smartphone or head-mounted display; and means for generating answers in response to questions from users and displaying them in AR or chat format. This allows users to receive information in real time from specific influencer AIs, enabling influencers to create new content based on user needs.

[0614] An "influencer" is an individual or group who influences a large number of followers through an online platform.

[0615] "Existing content" refers to digital information such as text, images, and videos that have already been created and published by influencers.

[0616] A "generative AI model" is an artificial intelligence that is trained using natural language processing and machine learning techniques based on collected data.

[0617] "Personal AI" is a generative AI model that is customized to reflect the style and expertise of a specific influencer.

[0618] An "inquiry" is a question or request for information that a user makes to the influencer AI.

[0619] A "smartphone" is a portable electronic device that, in addition to the functions of a mobile phone, has advanced computing power and Internet access capabilities.

[0620] A "head-mounted display" is a display device worn on the user's head, typically for displaying visual and auditory content.

[0621] "Real-time" refers to processing and responding to information and data instantly, without delay.

[0622] "Chat format" refers to a format in which a conversation is carried out between the user and the system by sending and receiving text messages.

[0623] "AR (Augmented Reality)" is a technology that displays digital information overlaid on images of the real world.

[0624] "Feedback" refers to reaction information based on comments, impressions, ratings, etc. provided by users.

[0625] The present invention is a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in chat or AR formats. The following describes an embodiment of the system in detail.

[0626] Server-side implementation

[0627] 1. Content collection and learning

[0628] The server uses APIs and web scraping technology to collect influencers' existing content. This content includes video data from video platforms and text and image data from social media. This data is stored in a database and, after preprocessing, is used as training data for the generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[0629] 2. Providing chat and AR responses

[0630] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer based on the inquiry, and the server sends the answer to the user's device. The answer is displayed on the user's device in chat format or AR format.

[0631] 3. Gathering and Summarizing Feedback

[0632] The server collects feedback from users, including their question history, comments, and impressions. It then uses a generative AI model to summarize the feedback and provide it to influencers, who can then create new content based on user needs.

[0633] Terminal side embodiment

[0634] 1. Use of the Application

[0635] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. If a head-mounted display is used, information is provided in AR format.

[0636] 2. Operations on the chat screen and AR screen

[0637] The questions entered by the user are sent from the device to the server. The responses received from the server are instantly displayed on the chat screen or AR screen, and the user can use the information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0638] 3. Providing Feedback

[0639] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[0640] User's embodiment

[0641] 1. Start a chat

[0642] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0643] 2. Receiving and using information

[0644] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0645] 3. Providing Feedback

[0646] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using a generative AI model, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[0647] Specific examples

[0648] If the user asks "Tell me about summer fashion," the following prompt will be generated:

[0649] "Users ask AI fashion influencers questions: Tell us about summer fashion"

[0650] This allows users to receive specific fashion advice in chat or AR format.

[0651] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0652] Step 1:

[0653] The server collects influencers' existing content. Specifically, it uses APIs and web scraping technology to collect text, image, and video data from video platforms and social media, and stores it in a database. The collected data undergoes preprocessing (noise removal, normalization, tokenization, etc.) and is prepared as training data for the generative AI model. The input is public content on the internet, and the output is preprocessed learning data.

[0654] Step 2:

[0655] The server uses the preprocessed data to train a generative AI model using natural language processing or machine learning algorithms (e.g., the GPT series). The input is the preprocessed content data, and the output is a generative AI model that has learned the style and expertise of a specific influencer. Specifically, the server splits the dataset and iteratively trains the model.

[0656] Step 3:

[0657] The server receives inquiries from users via a REST API. For example, if a user asks, "Tell me about summer fashion coordination," the server converts the inquiry into an appropriate format. The input is the user's inquiry text, and the output is data that has been appropriately analyzed from the inquiry. Specifically, the server uses natural language processing to tokenize and analyze the inquiry.

[0658] Step 4:

[0659] The server selects the optimal personal AI based on the content of the inquiry. This generates a response based on the style and expertise of a specific influencer. The input is the analyzed inquiry content and information from multiple personal AIs, and the output is the optimal personal AI. Specifically, it compares the characteristics of the inquiry content and the personal AIs and selects the AI ​​with the highest degree of match.

[0660] Step 5:

[0661] The server generates a response to the inquiry using the selected personal AI. The generated answer is sent to the user. The input is the inquiry content and the selected personal AI, and the output is the generated answer. Specifically, a prompt sentence is input into the generative AI model to obtain a natural conversational response.

[0662] Step 6:

[0663] The device displays the answers received from the server on the chat screen or in AR format. The input is the answer data sent from the server, and the output is the information visually displayed to the user. Specifically, the information is displayed in text format or augmented reality format on the screen of the chat application or AR display.

[0664] Step 7:

[0665] The user inputs feedback on the provided information. The input is the user's feedback text, and the output is the feedback data sent to the server. Specifically, the information is entered and sent from a feedback input form in the chat application.

[0666] Step 8:

[0667] The server collects feedback from users and summarizes it using a generative AI model. The input is the feedback data sent by the user, and the output is the summarized feedback information. Specifically, the server analyzes the feedback data using natural language processing technology, extracts key points, and generates a summary.

[0668] Step 9:

[0669] The server provides summarized feedback to influencers. The input is the summarized feedback data, and the output is the data provided in the form of notifications or reports to the influencers. Specific actions include sending the feedback to the influencers' dashboards or notification systems.

[0670] The above are the specific processing steps and their contents of the system that realizes the application example.

[0671] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0672] The present invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, provides information to users in a chat format, and combines it with an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions. The following describes an embodiment of the system in detail.

[0673] Server-side implementation

[0674] 1. Content collection and learning

[0675] The server first collects the influencer's existing content through APIs and web scraping, storing it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is then preprocessed and fed into a generative AI model. The generative AI model then learns from this data and generates a personal AI that reflects the style and expertise of the specific influencer.

[0676] 2. Providing chat-style responses

[0677] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[0678] 3. Emotion engine integration

[0679] The server is equipped with an emotion engine that analyzes user inquiries and feedback to recognize user emotions, such as joy, anger, sadness, and surprise, based on the context of the text.

[0680] 4. Emotion-Based Response Modulation

[0681] The server adjusts the personal AI's response based on the user's emotional information analyzed by the emotion engine. For example, if a user gives negative feedback such as "This idea was completely useless," the emotion engine will determine this as "anger," and the personal AI will adjust its response to be more polite, such as providing additional advice or citing past success stories.

[0682] 5. Collecting and Summarizing Feedback

[0683] The server collects user feedback and uses generative AI to summarize it. The summary, including the sentiment data of the feedback, is provided to influencers in a concise format, allowing them to create new content based on user needs.

[0684] Terminal side embodiment

[0685] 1. Use of the Application

[0686] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[0687] 2. Operations on the chat screen

[0688] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0689] 3. Providing Feedback

[0690] The feedback provided by the user is transmitted from the device to the server. This feedback includes new questions, comments, satisfaction ratings, and even emotional data. The device transmits this information in an appropriate format to facilitate analysis on the server side.

[0691] User's embodiment

[0692] 1. Start a chat

[0693] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0694] 2. Receiving and using information

[0695] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0696] 3. Receiving emotional responses

[0697] Based on the user's input query and feedback, the system uses an emotion engine to generate an appropriate response. For example, if the user is confused, the system will provide additional explanation.

[0698] 4. Providing Feedback

[0699] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server then collects this feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows influencers to more accurately understand users' needs and emotions and improve their content.

[0700] This will streamline communication between influencers and users, and promote the circulation of more personalized and sustainable content by taking emotions into consideration.

[0701] The processing flow will be explained below.

[0702] Server-side processing

[0703] Step 1:

[0704] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[0705] Step 2:

[0706] The server pre-processes the collected content, specifically cleansing, tokenizing, and batching the text data to prepare it for ingestion into the generative AI model.

[0707] Step 3:

[0708] The server trains a generative AI model, using the pre-processed content as training data and running feedforward and backpropagation on the model to generate a personal AI that reflects the style and expertise of a particular influencer.

[0709] Step 4:

[0710] The server receives inquiries from users via the REST API. For example, the user might ask, "Tell me about summer fashion coordination."

[0711] Step 5:

[0712] The server uses an emotion engine to analyze the content of the user's inquiry and recognize their emotion. For example, it can determine their interest from the context of "Please tell me."

[0713] Step 6:

[0714] The server selects an appropriate personal AI based on the user's inquiry and emotional information. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[0715] Step 7:

[0716] The selected personal AI generates an answer based on the user's query and emotional information, and the generated text is sent to the server to be prepared as the final answer.

[0717] Step 8:

[0718] The server then sends the generated answer to the user's device, such as "A cool linen shirt and denim skirt are recommended for this summer."

[0719] Step 9:

[0720] The server collects feedback from users. The feedback is received in text format, and the emotion engine simultaneously analyzes the emotion data. For example, feedback such as "This was helpful" is recognized as "satisfied / positive."

[0721] Step 10:

[0722] The server summarizes the collected feedback and its sentiment data, then uses generative AI to summarize it and provide it to influencers in a concise format, allowing them to create new content that reflects user needs and sentiment.

[0723] Terminal side processing

[0724] Step 1:

[0725] The terminal installs a system-specific application.

[0726] Step 2:

[0727] The user launches the application and logs in by entering their account information.

[0728] Step 3:

[0729] Once the application has logged in, it will display the chat screen.

[0730] Step 4:

[0731] Questions and requests entered by the user are sent from the device to the server.

[0732] Step 5:

[0733] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[0734] Step 6:

[0735] When the user provides feedback, the application sends the feedback to the server, along with the user's emotional data.

[0736] User Action

[0737] Step 1:

[0738] The user launches the application and logs in.

[0739] Step 2:

[0740] Users go to the chat screen and type in their questions or requests.

[0741] Step 3:

[0742] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[0743] Step 4:

[0744] When users provide feedback, they type their thoughts about the answer or follow-up questions into the chat window, and emotional data is automatically analyzed and sent along with the feedback.

[0745] Step 5:

[0746] Send feedback and have it delivered to the server, where it will be analyzed and summarized by the system and provided to the influencer.

[0747] This will enable more efficient communication between influencers and users, and by taking emotions into consideration, will enable a more personalized and sustainable circulation of content.

[0748] Example 2

[0749] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0750] In the conventional system, information provided through influencer content was one-way, making it difficult to respond flexibly to user emotions. Furthermore, there was no system in place to efficiently collect and analyze user feedback, making it difficult for influencers to accurately understand users' needs and emotions. This limited the user experience and prevented effective communication between influencers and users.

[0751] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0752] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the users; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for preprocessing text data; means for analyzing emotions that recognize the user's emotions; and means for adjusting response content based on the recognized emotions. This enables flexible information provision based on user emotions and efficient collection and analysis of feedback.

[0753] An "influencer" is an individual or organization that uses social media or other digital platforms to influence a large number of followers.

[0754] "Existing content" refers to digital media data such as videos, text, and images that have already been published by influencers.

[0755] "Collection Methods" refers to the technical means used to obtain existing content from influencers, using technologies such as APIs and web scraping.

[0756] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of data and generates natural language and images.

[0757] "Training method" refers to the technical process or method for training a generative AI model based on existing content.

[0758] "Personal AI" refers to artificial intelligence that learns the style and expertise of a specific influencer and provides information tailored to that influencer.

[0759] "Means for receiving inquiries" refers to the technical means by which the server receives inquiries from users, such as through a REST API.

[0760] "Means for selecting the optimal personal AI" refers to algorithms and technologies for selecting the appropriate personal AI based on the content of the user's inquiry.

[0761] "Means for sending answers" refers to the communication means for sending answers generated by personal AI to the user's device.

[0762] "Feedback Collection Means" refers to the technical means by which the Server collects comments and ratings provided by Users.

[0763] "Means for summarizing and providing feedback" refers to technology for summarizing collected feedback using a generative AI model and providing that summary to influencers.

[0764] "Means for preprocessing text data" refers to the technical means for analyzing and cleansing collected text data and converting it into a form that is easy for generative AI models to use.

[0765] "Sentiment analysis methods" refer to algorithms and technologies used to recognize and classify emotions from user text data.

[0766] "Means for adjusting response content" refers to technical means for appropriately changing the responses of personal AI based on the results of emotion analysis.

[0767] This invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, and provides information to users in chat format, as well as a system that combines an emotion engine that recognizes the user's emotions and adjusts responses based on them.

[0768] Server-side implementation

[0769] 1. Content collection and learning

[0770] The server first collects the influencer's existing content through APIs and web scraping. For example, it obtains video data from video platforms and collects text and image data from social networking services. This data is preprocessed and fed into the generative AI model. Preprocessing includes converting the video's audio to text, extracting features from the image data, and cleansing the text. The generative AI model then learns from this data to generate a personal AI that reflects the style and expertise of a specific influencer. For example, it obtains video data using YouTube's API, extracts audio from the video using FFmpeg, and converts it into text using the Google Cloud Speech-to-Text API. The collected text data is then formatted using natural language processing technology.

[0771] 2. Receiving and Processing User Requests

[0772] When a user makes a request through the application, such as "Tell me about summer fashion coordination," the request is sent to the server via a REST API. The server selects an appropriate personal AI and communicates the inquiry to that personal AI. The generative AI model generates an answer in response to the inquiry. If the answer is "I recommend a linen shirt and a denim skirt," the server sends this answer to the user's device.

[0773] 3. Emotion engine integration and emotion analysis

[0774] The server is integrated with an emotion engine that analyzes user inquiries and feedback to recognize emotions. For example, if a user sends feedback such as "This idea was completely useless," the emotion engine analyzes the text and determines that the emotion is "anger." This is done using natural language processing technology and emotion classification algorithms.

[0775] 4. Adjusting the response

[0776] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We apologize for the inconvenience. What do you think of these suggestions?" and sends it to the user.

[0777] 5. Collecting and Summarizing Feedback

[0778] The server collects user feedback and uses generative AI to summarize it, including emotional information, and provides it to influencers, who can then create new content based on user needs.

[0779] Terminal side embodiment

[0780] 1. Use of the Application

[0781] Users install the dedicated application and log in. After logging in, a chat screen appears, allowing users to freely post questions or requests.

[0782] 2. Operations on the chat screen

[0783] When a user enters a question, the content is sent from the device to the server. The server's response is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0784] 3. Providing Feedback

[0785] Users can enter feedback on the information provided and send it from their device to the server. This feedback can include new questions, comments, satisfaction ratings, and sentiment data. The device then sends this information in an appropriate format to facilitate analysis by the server.

[0786] User's embodiment

[0787] 1. Start a chat

[0788] Users can navigate the app to start a chat to request information from a specific influencer or find information related to a specific topic, for example, by typing, "Tell me about summer fashion."

[0789] 2. Receiving and using information

[0790] The system receives the personal AI's response from the server and utilizes that information. For example, it receives specific advice such as, "Recommended fashion items for summer are a linen shirt and a denim skirt."

[0791] 3. Receiving emotional responses

[0792] When a user enters a query or feedback, the emotion engine recognizes the user's emotions and generates an appropriate response. For example, if the user expresses dissatisfaction by saying "this advice was not helpful," the engine will provide additional advice.

[0793] 4. Providing Feedback

[0794] Users can input their thoughts on the information provided or any follow-up questions as feedback and provide it to the server. The server then collects the feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows the influencer to more accurately understand the needs and sentiments of users and improve their content.

[0795] Examples of prompt statements

[0796] "Please tell me about your summer fashion coordination."

[0797] "This idea was completely useless."

[0798] "Please tell me some more casual outfit ideas."

[0799] This will streamline communication between influencers and users, and by taking emotions into consideration, more personalized information will be provided.

[0800] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0801] Step 1:

[0802] Content Collection

[0803] The server collects influencers' existing content through APIs and web scraping. Specifically, it uses the YouTube API to obtain video data and social media APIs to collect text and image data. The server specifies the influencer's account or specific keywords as input, and obtains video data, text data, and image data as output.

[0804] What it does: The server periodically sends a YouTube API request to get the metadata and video URL of new videos, then uses another API request to download the videos.

[0805] Step 2:

[0806] Data Preprocessing

[0807] The server preprocesses the acquired data. Specifically, it converts the video audio into text (e.g., using FFmpeg), extracts features from the image data (e.g., using TensorFlow), and cleanses the text data (e.g., removing unnecessary characters and tags using regular expressions). It receives video, text, and image data as input and generates preprocessed data as output.

[0808] What it does: It converts the video into an audio file using FFmpeg, then converts the audio to text using the Google Cloud Speech-to-Text API, and then formats the text using a natural language processing library.

[0809] Step 3:

[0810] Training generative AI models

[0811] The server uses the preprocessed data to train a generative AI model. Specifically, it uses GPT-3, for example, to learn style and expertise from influencer content. It receives the preprocessed data as input and generates a generative AI model as output.

[0812] Specific operation: The server uses TensorFlow or PyTorch to train a generative AI model and saves the model after learning is complete.

[0813] Step 4:

[0814] Receiving a user request

[0815] The server receives user inquiries via a REST API. For example, it receives a request such as, "Tell me about summer fashion coordination." It receives the user's question as input and sends the request to the appropriate personal AI as output.

[0816] What happens: The user's request is sent to the server as an HTTP POST request and parsed.

[0817] Step 5:

[0818] Generate and send a response

[0819] The server passes the received request to the generative AI model, which then generates an appropriate answer. For example, it might generate an answer such as "I recommend a linen shirt and a denim skirt," and the server sends that answer to the user's device. The generative AI model receives the user request and the generated answer as input and sends the generated answer to the user as output.

[0820] How it works: The generative AI model generates a response, and the server displays the response on the user's chat screen in real time.

[0821] Step 6:

[0822] Emotion analysis

[0823] The server uses an emotion engine to analyze the text of user inquiries and feedback and recognize the user's emotions. For example, if the user gives negative feedback such as "This idea was completely useless," it will identify it as "anger." It receives user feedback as input and generates emotion data as output.

[0824] Specific operation: Extracts emotions from text using a sentiment analysis algorithm and assigns sentiment tags.

[0825] Step 7:

[0826] Regulating responses based on emotions

[0827] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We're sorry for the inconvenience. What do you think of these suggestions?" and sends it to the user. It receives the sentiment data and the original answer as input and generates the adjusted response as output.

[0828] Specific operation: Based on the emotion data, a new prompt is given to the generative AI model to generate an appropriate response.

[0829] Step 8:

[0830] Collecting and summarizing feedback

[0831] The server collects user feedback and uses generative AI to summarize it. The feedback includes emotional information and provides it to influencers, who can then create new content based on user needs. The server receives user feedback as input and generates summarized feedback as output.

[0832] Specific operation: Collect multiple feedback data, generate a summary using natural language processing technology, and notify the influencer of the summary.

[0833] (Application example 2)

[0834] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0835] In recent years, as the influence of influencers has grown, there has been an increasing demand for information provision systems that utilize influencer content. However, existing systems have difficulty generating appropriate responses that take user emotions into account, limiting the improvement of user experience. There is also a growing need for systems that not only generate personal AI that reflects the style of a specific influencer, but also have the ability to adjust responses according to different emotional states. Furthermore, there is a need for a mechanism that allows users to receive more personalized information in virtual stores. The present invention aims to solve these problems and provide a high-quality information provision system that reflects user emotions.

[0836] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting existing content of influencers, means for training a generative AI model based on the collected existing content, means for generating a personal AI for each influencer using the generative AI model, means for receiving inquiries from users, means for selecting an optimal personal AI based on the inquiries, means for sending answers generated by the personal AI to the user, means for collecting feedback from users, means for summarizing the feedback and providing it to the influencer, means for analyzing the user's emotions, and means for adjusting the response of the personal AI based on the analyzed emotions. This makes it possible to adjust responses according to the user's emotional state, thereby enabling the provision of more personalized, high-quality information.

[0837] An "influencer" is someone who has a large following and whose opinions and actions influence other people.

[0838] "Existing content" refers to data such as text, images, and videos that influencers have previously posted.

[0839] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task based on collected data.

[0840] "Personal AI" refers to artificial intelligence generated to reflect the style and expertise of a specific influencer.

[0841] "Enquiry" means a question or request that a user enters into a system seeking information.

[0842] A "response" is an answer that a personal AI provides to a user's inquiry.

[0843] "Feedback" refers to the evaluations and opinions that users give in response to suggestions and advice.

[0844] "Sentiment analysis" is the process of analyzing and identifying a user's emotional state from their text and behavior.

[0845] "Response adjustment" is the process of appropriately modifying the answers provided by personal AI based on analyzed emotional data.

[0846] This invention is a system that collects existing influencer content, creates a generative AI model, and generates a personal AI for each influencer, which provides appropriate responses to user inquiries. It also analyzes user emotions and adjusts responses based on that information, thereby providing more personalized information.

[0847] Server-side implementation

[0848] The server first collects existing content from influencers through APIs and web scraping. The collected data is stored in a database, where it undergoes pre-processing before being fed into a generative AI model. This generative AI model learns the style and expertise of a specific influencer and generates a personal AI based on the results.

[0849] Inquiries from users are received via REST APIs, and the most suitable personal AI is selected based on the content of the inquiry. The selected personal AI generates a response to the inquiry and sends it from the server to the user's device. At this time, the server is equipped with an emotion analysis engine that analyzes the user's emotions and adjusts the response based on that information.

[0850] By utilizing the sentiment analysis engine, if a user sends negative feedback, such as "This advice was not helpful," the sentiment engine will detect "anger" and the personal AI will adjust its response to be more polite. This response adjustment process may include providing additional advice or citing past success stories.

[0851] Terminal side embodiment

[0852] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. Questions entered by the user are sent from the device to the server, and responses received from the server are immediately displayed on the chat screen.

[0853] The feedback sent from the device includes satisfaction ratings and emotional data. This information is sent to a server for analysis. The emotional data is fed back into the generative AI model, helping to improve the accuracy of each influencer's personal AI.

[0854] User's embodiment

[0855] Users operate the application and start a chat when they want information from a specific influencer. They receive a response to their input from a personal AI, which then utilizes that information. For example, if a user asks, "I'd like some advice on casual summer style," they can receive specific advice from the influencer AI, such as, "I recommend a linen shirt and a denim skirt."

[0856] The feedback provided by users includes information based on sentiment analysis. For example, if a user is confused and says, "I want more specific suggestions," the system will provide additional, detailed advice. This feedback is collected on the server and provided to influencers, allowing them to generate new content based on user needs.

[0857] Examples and prompts

[0858] As a concrete example, imagine a scenario in which a user is choosing fashion in a virtual store. For example, if the user asks, "I'd like some advice on a casual summer style," the system will respond, "I recommend a linen shirt and a denim skirt."

[0859] Example prompt sentence:

[0860] "A user is asking about casual summer style. What suggestions can you offer?"

[0861] "The user is confused. Please provide more specific advice."

[0862] This allows them to take into account the user's emotions and provide a more personalized response.

[0863] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0864] Step 1:

[0865] The server collects existing content from influencers through APIs and web scraping, and stores it in a database in the form of video, text, images, etc.

[0866] Input: Influencer content (e.g., video, text, images)

[0867] Output: Content data stored in the database

[0868] How it works: The server uses web scraping tools and APIs to retrieve content from social media and video platforms and stores it in a database.

[0869] Step 2:

[0870] The server preprocesses the collected data and feeds it into the generative AI model. Data preprocessing includes removing unnecessary information and standardizing the format.

[0871] Input: Content data in the database

[0872] Output: Preprocessed data

[0873] Specific operation: The server organizes the data using data cleaning techniques and performs preprocessing such as tokenizing text and resizing images.

[0874] Step 3:

[0875] The server uses the preprocessed data to train a generative AI model, which acts as a personal AI.

[0876] Input: Preprocessed data

[0877] Output: A trained generative AI model

[0878] How it works: The server runs machine learning algorithms and trains generative AI models based on the collected data.

[0879] Step 4:

[0880] The user's device starts the application and displays the login screen. The user logs in.

[0881] Input: User login information

[0882] Output: User authentication result

[0883] Specific operation: The terminal sends the login information entered by the user to the server, and the server performs authentication.

[0884] Step 5:

[0885] The terminal displays a chat screen and receives inquiries from users.

[0886] Input: User's inquiry

[0887] Output: Inquiry content displayed on the chat screen

[0888] Specific behavior: Sends the text entered by the user on the chat screen to the server.

[0889] Step 6:

[0890] The server receives the user's inquiry and selects the most suitable personal AI.

[0891] Input: User's inquiry

[0892] Output: Selected personal AI

[0893] Specific operation: The server analyzes the query content and selects the most suitable personal AI model.

[0894] Step 7:

[0895] The server sends the response generated by the selected personal AI to the user's device.

[0896] Input: Enquiry details, personal AI model

[0897] Output: Response by personal AI

[0898] Specific operation: The personal AI generates a response based on the inquiry, and the server sends that response to the user's device.

[0899] Step 8:

[0900] The server receives the user's feedback and analyzes its content and sentiment.

[0901] Input: User feedback

[0902] Output: Parsed emotion data

[0903] Specific operation: The emotion analysis engine built into the server analyzes the feedback text to determine the emotion.

[0904] Step 9:

[0905] The server adjusts the personal AI's responses based on the analyzed emotions.

[0906] Input: Emotional data, Personal AI response

[0907] Output: Adjusted response

[0908] Specific behavior: The server takes into account the emotional data and appropriately modifies the personal AI's responses, for example, making the response more polite if anger is detected.

[0909] Step 10:

[0910] The server summarizes the user's feedback and stores it for providing to the influencer.

[0911] Input: User feedback

[0912] Output: Summarized feedback data

[0913] Specific operation: The server uses generative AI to summarize the feedback content and saves the results in a database.

[0914] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0915] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0916] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0917] [Third embodiment]

[0918] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0919] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0920] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0921] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0922] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0923] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0924] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0925] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0926] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0927] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0928] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0929] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0930] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[0931] Server-side implementation

[0932] 1. Content collection and learning

[0933] The server collects influencers' existing content through APIs and web scraping and stores it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is preprocessed and fed into a generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[0934] 2. Providing chat-style responses

[0935] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[0936] 3. Gathering and Summarizing Feedback

[0937] The server collects feedback from users, including their question history, comments, and impressions. Generative AI is used to summarize the feedback and provide it to influencers in a concise format, allowing the influencers to create new content based on user needs.

[0938] Terminal side embodiment

[0939] 1. Use of the Application

[0940] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[0941] 2. Operations on the chat screen

[0942] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[0943] 3. Providing Feedback

[0944] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[0945] User's embodiment

[0946] 1. Start a chat

[0947] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[0948] 2. Receiving and using information

[0949] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[0950] 3. Providing Feedback

[0951] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using generative AI, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[0952] This will streamline communication between influencers and users and promote the circulation of sustainable content.

[0953] The processing flow will be explained below.

[0954] Server-side processing

[0955] Step 1:

[0956] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[0957] Step 2:

[0958] The server pre-processes the collected content and feeds it into the generative AI model. Pre-processing includes text cleansing (noise removal), tokenization (breaking down words and sentences), and data batching.

[0959] Step 3:

[0960] The server trains a generative AI model to generate a personal AI that reflects the style and expertise of a specific influencer by running feedforward and backpropagation on the model using a large dataset.

[0961] Step 4:

[0962] The server receives inquiries from users via REST APIs, etc. The inquiry content is sent to the server in text format.

[0963] Step 5:

[0964] The server selects an appropriate personal AI based on the user's inquiry. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[0965] Step 6:

[0966] The selected personal AI generates a response to the query, and the generated text is sent to the server and prepared as the final answer.

[0967] Step 7:

[0968] The server then sends the generated answer to the user's device, where it is displayed in the user's application.

[0969] Step 8:

[0970] The server collects feedback from users, which is received in text format.

[0971] Step 9:

[0972] The server uses generative AI to summarize the collected feedback, which involves briefly summarizing the feedback and categorizing positive comments and requests for improvement.

[0973] Step 10:

[0974] The server provides summarized feedback to the influencer, who then uses it to plan and create new content.

[0975] Terminal side processing

[0976] Step 1:

[0977] A system-specific application is installed on the user's device.

[0978] Step 2:

[0979] The user launches the application and logs in by entering their account information.

[0980] Step 3:

[0981] Once the application has logged in, it will display the chat screen.

[0982] Step 4:

[0983] Questions and requests entered by the user are sent from the application to the server.

[0984] Step 5:

[0985] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[0986] Step 6:

[0987] If the user provides feedback, the application sends the feedback to the server.

[0988] User Action

[0989] Step 1:

[0990] The user launches the application and logs in.

[0991] Step 2:

[0992] Users go to the chat screen and type in their questions or requests.

[0993] Step 3:

[0994] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[0995] Step 4:

[0996] When users provide feedback, they type their thoughts about the answer to their question or any follow-up questions into the chat screen.

[0997] Step 5:

[0998] Send feedback and have it delivered to the server.

[0999] This will enable efficient communication between influencers and users, realizing a sustainable circulation of content.

[1000] Example 1

[1001] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1002] Currently, there is a lack of systems that can efficiently collect influencer content and provide personalized information to users. There is also a lack of efficient methods for conveying user feedback to influencers. This makes communication between users and influencers difficult, making it difficult to create new content based on user needs.

[1003] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1004] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiry; means for sending responses generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for storing the collected content in a database; means for using natural language processing technology to preprocess the collected data; means for inputting the preprocessed data as training data into the generative AI model; means for executing the generative AI model to generate responses; and means for summarizing the collected feedback using natural language processing technology. This not only effectively collects existing content from influencers, generates a personal AI based on the collected content, and provides personalized responses to users, but also efficiently communicates user feedback to influencers, facilitating the creation of new content.

[1005] An "influencer" is a person who has influence mainly on social media and online platforms and disseminates their opinions and information to a large number of followers.

[1006] "Content" is a general term for information, including data formats such as text, images, and videos, that are produced and disseminated by influencers.

[1007] A "generative AI model" is an artificial intelligence technology that uses machine learning algorithms to learn from large amounts of data and generate text that is as natural as a human's.

[1008] "Personal AI" is a generative AI model that learns the style and expertise of a specific influencer and generates information similar to that influencer.

[1009] A "query" is a question or request that a user enters into a system to ask for specific information.

[1010] "Feedback" refers to the user's reaction to the information provided, such as rating, comments, or additional questions.

[1011] A "database" is a system for efficiently storing, managing, and searching collected data.

[1012] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, including text cleaning, tokenization, and labeling.

[1013] "Training data" is a pre-processed data set used to train a generative AI model.

[1014] A "REST API" is a type of web service that is an interface for exchanging information between a server and a client using the HTTP protocol.

[1015] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[1016] Server-side implementation

[1017] The server uses APIs and web scraping techniques to collect existing content from influencers. Specifically, it uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The collected data is then stored in a NoSQL database such as MongoDB.

[1018] The server then preprocesses the collected data using natural language processing techniques, such as NLP libraries like SpaCy and NLTK. Preprocessing involves cleaning, tokenizing, and labeling the text. The video data is then converted to text using speech recognition techniques (e.g., Google Cloud Speech-to-Text API).

[1019] The preprocessed data is then used to train a generative AI model, such as GPT-3 or BERT, using machine learning frameworks like PyTorch or TensorFlow to generate a personalized AI model that reflects the style and expertise of a specific influencer.

[1020] The server receives user queries via a REST API. Frameworks such as Flask and FastAPI are used for this. If a user asks, "Tell me about summer fashion coordination," the request is sent to the personal AI, which generates a response. The response is sent back to the server, which then sends it to the user's device.

[1021] In addition, the server collects feedback from users, which is summarized using NLP technology and stored in a database. This summarized feedback is provided to influencers to help them create new content based on user needs.

[1022] Terminal side embodiment

[1023] A mobile application dedicated to this system is installed on the user's device. This application was developed using React Native and Flutter. The user launches the app and logs in. Once authentication is successful, a chat screen appears, allowing the user to freely enter questions or requests.

[1024] When a user types a question into the chat screen, the data is sent from the device to the server. For example, a user might type, "What are the latest summer fashion trends?" and press the send button. This request is sent to the server, which then forwards it to the appropriate personal AI. The generated response is sent from the server to the device and displayed on the chat screen. For example, specific advice such as, "I recommend a linen shirt and a denim skirt" is displayed.

[1025] Users can input their thoughts about the information provided and any follow-up questions as feedback, which is then sent from the device to the server. This feedback also includes a satisfaction rating. The device then sends this information in an appropriate format, making it easier for the server to analyze.

[1026] User's embodiment

[1027] Users interact with the app and initiate a chat to request information from a specific influencer or to find information related to a specific topic, for example, by typing in a question like "summer fashion" and waiting for a response to their question.

[1028] The system receives the answers sent by the influencer AI from the server and utilizes that information. For example, you can receive specific fashion advice such as "Linen shirts and denim skirts are trending."

[1029] Finally, users can provide feedback to the system, including their thoughts and ratings on the information provided. This feedback is then used by the system and influencers to create more optimized content.

[1030] Prompt Sentence Examples

[1031] As a concrete example, suppose a user enters the following question into an application:

[1032] "Please tell me about your summer fashion coordination."

[1033] An example prompt sentence to generate a response to this request:

[1034] "Provide summer fashion recommendations. Users want casual, comfortable outfits."

[1035] By inputting this prompt into a generative AI model, the influencer AI can provide specific suggestions to users, such as "I recommend a linen shirt, a denim skirt, and lightweight sandals."

[1036] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1037] System program processing flow and specific operations

[1038] Step 1:

[1039] The server collects existing content from influencers. It uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The input is the website URL and YouTube video ID, and the output is the extracted text data and downloaded video data. The collected data is stored in MongoDB.

[1040] Step 2:

[1041] The server preprocesses the collected data. It uses an NLP library (e.g., SpaCy or NLTK) to clean, tokenize, and label the text. It then uses the Google Cloud Speech-to-Text API to extract audio from the video data and convert it to text. The input is the raw data collected in step 1, and the output is the preprocessed text data.

[1042] Step 3:

[1043] The server inputs the preprocessed data into a generative AI model and trains the model. For example, it uses PyTorch or TensorFlow to train a model like GPT-3 or BERT. The input is preprocessed text data, and the output is a personal AI model that reflects the style and expertise of a specific influencer.

[1044] Step 4:

[1045] A user launches the application on their device and logs in. The app is developed using React Native and Flutter. If user authentication is successful, a chat screen is displayed. The input is the user's authentication information, and the output is the success / failure status of the authentication and the display of the chat screen.

[1046] Step 5:

[1047] The user enters a question or request on the chat screen, and the device sends the data to the server. For example, a question might be, "Tell me about summer fashion coordination." The input is the text of the question entered by the user, and the output is the request sent to the server.

[1048] Step 6:

[1049] The server sends the received user inquiry to the appropriate personal AI and generates a response. The generated response is received again by the server and sent to the user's device. For example, the personal AI generates an answer such as "I recommend a linen shirt and a denim skirt." The input is the text of the user's inquiry, and the output is the text of the response generated by the personal AI.

[1050] Step 7:

[1051] The terminal receives the response sent from the server and displays it on the chat screen. The user can refer to the displayed information. For example, the advice "We recommend a linen shirt and a denim skirt" is displayed. The input is the text of the response received from the server, and the output is to display that text on the chat screen.

[1052] Step 8:

[1053] The user inputs feedback on the provided information and sends it to the server from the terminal. The feedback includes ratings and follow-up questions. The input is the text of the feedback entered by the user, and the output is the feedback sent to the server.

[1054] Step 9:

[1055] The server summarizes the collected feedback using natural language processing technology and stores it in a database. The summarized feedback is provided to influencers. The input is the text of the feedback received from users, and the output is the summarized text of the feedback. This allows influencers to use it to create new content.

[1056] This will facilitate smooth communication between users and influencers, making it possible to create new content based on user needs.

[1057] (Application example 1)

[1058] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1059] In recent years, there has been an increasing demand for marketing and content provision using influencers. However, there are limited ways for users to receive fast, personalized information from specific influencers, making it difficult for influencers to accurately understand user needs and create new content. Furthermore, existing content provided in the form of videos, images, and text makes it difficult for users to obtain information in real time and lacks interactivity. A new system is needed to solve this issue and streamline communication between users and influencers.

[1060] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1061] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for conducting real-time dialogue via a smartphone or head-mounted display; and means for generating answers in response to questions from users and displaying them in AR or chat format. This allows users to receive information in real time from specific influencer AIs, enabling influencers to create new content based on user needs.

[1062] An "influencer" is an individual or group who influences a large number of followers through an online platform.

[1063] "Existing content" refers to digital information such as text, images, and videos that have already been created and published by influencers.

[1064] A "generative AI model" is an artificial intelligence that is trained using natural language processing and machine learning techniques based on collected data.

[1065] "Personal AI" is a generative AI model that is customized to reflect the style and expertise of a specific influencer.

[1066] An "inquiry" is a question or request for information that a user makes to the influencer AI.

[1067] A "smartphone" is a portable electronic device that, in addition to the functions of a mobile phone, has advanced computing power and Internet access capabilities.

[1068] A "head-mounted display" is a display device worn on the user's head, typically for displaying visual and auditory content.

[1069] "Real-time" refers to processing and responding to information and data instantly, without delay.

[1070] "Chat format" refers to a format in which a conversation is carried out between the user and the system by sending and receiving text messages.

[1071] "AR (Augmented Reality)" is a technology that displays digital information overlaid on images of the real world.

[1072] "Feedback" refers to reaction information based on comments, impressions, ratings, etc. provided by users.

[1073] The present invention is a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in chat or AR formats. The following describes an embodiment of the system in detail.

[1074] Server-side implementation

[1075] 1. Content collection and learning

[1076] The server uses APIs and web scraping technology to collect influencers' existing content. This content includes video data from video platforms and text and image data from social media. This data is stored in a database and, after preprocessing, is used as training data for the generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[1077] 2. Providing chat and AR responses

[1078] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer based on the inquiry, and the server sends the answer to the user's device. The answer is displayed on the user's device in chat format or AR format.

[1079] 3. Gathering and Summarizing Feedback

[1080] The server collects feedback from users, including their question history, comments, and impressions. It then uses a generative AI model to summarize the feedback and provide it to influencers, who can then create new content based on user needs.

[1081] Terminal side embodiment

[1082] 1. Use of the Application

[1083] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. If a head-mounted display is used, information is provided in AR format.

[1084] 2. Operations on the chat screen and AR screen

[1085] The questions entered by the user are sent from the device to the server. The responses received from the server are instantly displayed on the chat screen or AR screen, and the user can use the information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1086] 3. Providing Feedback

[1087] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[1088] User's embodiment

[1089] 1. Start a chat

[1090] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[1091] 2. Receiving and using information

[1092] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[1093] 3. Providing Feedback

[1094] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using a generative AI model, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[1095] Specific examples

[1096] If the user asks "Tell me about summer fashion," the following prompt will be generated:

[1097] "Users ask AI fashion influencers questions: Tell us about summer fashion"

[1098] This allows users to receive specific fashion advice in chat or AR format.

[1099] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1100] Step 1:

[1101] The server collects influencers' existing content. Specifically, it uses APIs and web scraping technology to collect text, image, and video data from video platforms and social media, and stores it in a database. The collected data undergoes preprocessing (noise removal, normalization, tokenization, etc.) and is prepared as training data for the generative AI model. The input is public content on the internet, and the output is preprocessed learning data.

[1102] Step 2:

[1103] The server uses the preprocessed data to train a generative AI model using natural language processing or machine learning algorithms (e.g., the GPT series). The input is the preprocessed content data, and the output is a generative AI model that has learned the style and expertise of a specific influencer. Specifically, the server splits the dataset and iteratively trains the model.

[1104] Step 3:

[1105] The server receives inquiries from users via a REST API. For example, if a user asks, "Tell me about summer fashion coordination," the server converts the inquiry into an appropriate format. The input is the user's inquiry text, and the output is data that has been appropriately analyzed from the inquiry. Specifically, the server uses natural language processing to tokenize and analyze the inquiry.

[1106] Step 4:

[1107] The server selects the optimal personal AI based on the content of the inquiry. This generates a response based on the style and expertise of a specific influencer. The input is the analyzed inquiry content and information from multiple personal AIs, and the output is the optimal personal AI. Specifically, it compares the characteristics of the inquiry content and the personal AIs and selects the AI ​​with the highest degree of match.

[1108] Step 5:

[1109] The server generates a response to the inquiry using the selected personal AI. The generated answer is sent to the user. The input is the inquiry content and the selected personal AI, and the output is the generated answer. Specifically, a prompt sentence is input into the generative AI model to obtain a natural conversational response.

[1110] Step 6:

[1111] The device displays the answers received from the server on the chat screen or in AR format. The input is the answer data sent from the server, and the output is the information visually displayed to the user. Specifically, the information is displayed in text format or augmented reality format on the screen of the chat application or AR display.

[1112] Step 7:

[1113] The user inputs feedback on the provided information. The input is the user's feedback text, and the output is the feedback data sent to the server. Specifically, the information is entered and sent from a feedback input form in the chat application.

[1114] Step 8:

[1115] The server collects feedback from users and summarizes it using a generative AI model. The input is the feedback data sent by the user, and the output is the summarized feedback information. Specifically, the server analyzes the feedback data using natural language processing technology, extracts key points, and generates a summary.

[1116] Step 9:

[1117] The server provides summarized feedback to influencers. The input is the summarized feedback data, and the output is the data provided in the form of notifications or reports to the influencers. Specific actions include sending the feedback to the influencers' dashboards or notification systems.

[1118] The above are the specific processing steps and their contents of the system that realizes the application example.

[1119] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1120] The present invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, provides information to users in a chat format, and combines it with an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions. The following describes an embodiment of the system in detail.

[1121] Server-side implementation

[1122] 1. Content collection and learning

[1123] The server first collects the influencer's existing content through APIs and web scraping, storing it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is then preprocessed and fed into a generative AI model. The generative AI model then learns from this data and generates a personal AI that reflects the style and expertise of the specific influencer.

[1124] 2. Providing chat-style responses

[1125] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[1126] 3. Emotion engine integration

[1127] The server is equipped with an emotion engine that analyzes user inquiries and feedback to recognize user emotions, such as joy, anger, sadness, and surprise, based on the context of the text.

[1128] 4. Emotion-Based Response Modulation

[1129] The server adjusts the personal AI's response based on the user's emotional information analyzed by the emotion engine. For example, if a user gives negative feedback such as "This idea was completely useless," the emotion engine will determine this as "anger," and the personal AI will adjust its response to be more polite, such as providing additional advice or citing past success stories.

[1130] 5. Collecting and Summarizing Feedback

[1131] The server collects user feedback and uses generative AI to summarize it. The summary, including the sentiment data of the feedback, is provided to influencers in a concise format, allowing them to create new content based on user needs.

[1132] Terminal side embodiment

[1133] 1. Use of the Application

[1134] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[1135] 2. Operations on the chat screen

[1136] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1137] 3. Providing Feedback

[1138] The feedback provided by the user is transmitted from the device to the server. This feedback includes new questions, comments, satisfaction ratings, and even emotional data. The device transmits this information in an appropriate format to facilitate analysis on the server side.

[1139] User's embodiment

[1140] 1. Start a chat

[1141] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[1142] 2. Receiving and using information

[1143] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[1144] 3. Receiving emotional responses

[1145] Based on the user's input query and feedback, the system uses an emotion engine to generate an appropriate response. For example, if the user is confused, the system will provide additional explanation.

[1146] 4. Providing Feedback

[1147] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server then collects this feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows influencers to more accurately understand users' needs and emotions and improve their content.

[1148] This will streamline communication between influencers and users, and promote the circulation of more personalized and sustainable content by taking emotions into consideration.

[1149] The processing flow will be explained below.

[1150] Server-side processing

[1151] Step 1:

[1152] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[1153] Step 2:

[1154] The server pre-processes the collected content, specifically cleansing, tokenizing, and batching the text data to prepare it for ingestion into the generative AI model.

[1155] Step 3:

[1156] The server trains a generative AI model, using the pre-processed content as training data and running feedforward and backpropagation on the model to generate a personal AI that reflects the style and expertise of a particular influencer.

[1157] Step 4:

[1158] The server receives inquiries from users via the REST API. For example, the user might ask, "Tell me about summer fashion coordination."

[1159] Step 5:

[1160] The server uses an emotion engine to analyze the content of the user's inquiry and recognize their emotion. For example, it can determine their interest from the context of "Please tell me."

[1161] Step 6:

[1162] The server selects an appropriate personal AI based on the user's inquiry and emotional information. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[1163] Step 7:

[1164] The selected personal AI generates an answer based on the user's query and emotional information, and the generated text is sent to the server to be prepared as the final answer.

[1165] Step 8:

[1166] The server then sends the generated answer to the user's device, such as "A cool linen shirt and denim skirt are recommended for this summer."

[1167] Step 9:

[1168] The server collects feedback from users. The feedback is received in text format, and the emotion engine simultaneously analyzes the emotion data. For example, feedback such as "This was helpful" is recognized as "satisfied / positive."

[1169] Step 10:

[1170] The server summarizes the collected feedback and its sentiment data, then uses generative AI to summarize it and provide it to influencers in a concise format, allowing them to create new content that reflects user needs and sentiment.

[1171] Terminal side processing

[1172] Step 1:

[1173] The terminal installs a system-specific application.

[1174] Step 2:

[1175] The user launches the application and logs in by entering their account information.

[1176] Step 3:

[1177] Once the application has logged in, it will display the chat screen.

[1178] Step 4:

[1179] Questions and requests entered by the user are sent from the device to the server.

[1180] Step 5:

[1181] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[1182] Step 6:

[1183] When the user provides feedback, the application sends the feedback to the server, along with the user's emotional data.

[1184] User Action

[1185] Step 1:

[1186] The user launches the application and logs in.

[1187] Step 2:

[1188] Users go to the chat screen and type in their questions or requests.

[1189] Step 3:

[1190] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[1191] Step 4:

[1192] When users provide feedback, they type their thoughts about the answer or follow-up questions into the chat window, and emotional data is automatically analyzed and sent along with the feedback.

[1193] Step 5:

[1194] Send feedback and have it delivered to the server, where it will be analyzed and summarized by the system and provided to the influencer.

[1195] This will enable more efficient communication between influencers and users, and by taking emotions into consideration, will enable a more personalized and sustainable circulation of content.

[1196] Example 2

[1197] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1198] In the conventional system, information provided through influencer content was one-way, making it difficult to respond flexibly to user emotions. Furthermore, there was no system in place to efficiently collect and analyze user feedback, making it difficult for influencers to accurately understand users' needs and emotions. This limited the user experience and prevented effective communication between influencers and users.

[1199] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1200] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the users; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for preprocessing text data; means for analyzing emotions that recognize the user's emotions; and means for adjusting response content based on the recognized emotions. This enables flexible information provision based on user emotions and efficient collection and analysis of feedback.

[1201] An "influencer" is an individual or organization that uses social media or other digital platforms to influence a large number of followers.

[1202] "Existing content" refers to digital media data such as videos, text, and images that have already been published by influencers.

[1203] "Collection Methods" refers to the technical means used to obtain existing content from influencers, using technologies such as APIs and web scraping.

[1204] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of data and generates natural language and images.

[1205] "Training method" refers to the technical process or method for training a generative AI model based on existing content.

[1206] "Personal AI" refers to artificial intelligence that learns the style and expertise of a specific influencer and provides information tailored to that influencer.

[1207] "Means for receiving inquiries" refers to the technical means by which the server receives inquiries from users, such as through a REST API.

[1208] "Means for selecting the optimal personal AI" refers to algorithms and technologies for selecting the appropriate personal AI based on the content of the user's inquiry.

[1209] "Means for sending answers" refers to the communication means for sending answers generated by personal AI to the user's device.

[1210] "Feedback Collection Means" refers to the technical means by which the Server collects comments and ratings provided by Users.

[1211] "Means for summarizing and providing feedback" refers to technology for summarizing collected feedback using a generative AI model and providing that summary to influencers.

[1212] "Means for preprocessing text data" refers to the technical means for analyzing and cleansing collected text data and converting it into a form that is easy for generative AI models to use.

[1213] "Sentiment analysis methods" refer to algorithms and technologies used to recognize and classify emotions from user text data.

[1214] "Means for adjusting response content" refers to technical means for appropriately changing the responses of personal AI based on the results of emotion analysis.

[1215] This invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, and provides information to users in chat format, as well as a system that combines an emotion engine that recognizes the user's emotions and adjusts responses based on them.

[1216] Server-side implementation

[1217] 1. Content collection and learning

[1218] The server first collects the influencer's existing content through APIs and web scraping. For example, it obtains video data from video platforms and collects text and image data from social networking services. This data is preprocessed and fed into the generative AI model. Preprocessing includes converting the video's audio to text, extracting features from the image data, and cleansing the text. The generative AI model then learns from this data to generate a personal AI that reflects the style and expertise of a specific influencer. For example, it obtains video data using YouTube's API, extracts audio from the video using FFmpeg, and converts it into text using the Google Cloud Speech-to-Text API. The collected text data is then formatted using natural language processing technology.

[1219] 2. Receiving and Processing User Requests

[1220] When a user makes a request through the application, such as "Tell me about summer fashion coordination," the request is sent to the server via a REST API. The server selects an appropriate personal AI and communicates the inquiry to that personal AI. The generative AI model generates an answer in response to the inquiry. If the answer is "I recommend a linen shirt and a denim skirt," the server sends this answer to the user's device.

[1221] 3. Emotion engine integration and emotion analysis

[1222] The server is integrated with an emotion engine that analyzes user inquiries and feedback to recognize emotions. For example, if a user sends feedback such as "This idea was completely useless," the emotion engine analyzes the text and determines that the emotion is "anger." This is done using natural language processing technology and emotion classification algorithms.

[1223] 4. Adjusting the response

[1224] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We apologize for the inconvenience. What do you think of these suggestions?" and sends it to the user.

[1225] 5. Collecting and Summarizing Feedback

[1226] The server collects user feedback and uses generative AI to summarize it, including emotional information, and provides it to influencers, who can then create new content based on user needs.

[1227] Terminal side embodiment

[1228] 1. Use of the Application

[1229] Users install the dedicated application and log in. After logging in, a chat screen appears, allowing users to freely post questions or requests.

[1230] 2. Operations on the chat screen

[1231] When a user enters a question, the content is sent from the device to the server. The server's response is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1232] 3. Providing Feedback

[1233] Users can enter feedback on the information provided and send it from their device to the server. This feedback can include new questions, comments, satisfaction ratings, and sentiment data. The device then sends this information in an appropriate format to facilitate analysis by the server.

[1234] User's embodiment

[1235] 1. Start a chat

[1236] Users can navigate the app to start a chat to request information from a specific influencer or find information related to a specific topic, for example, by typing, "Tell me about summer fashion."

[1237] 2. Receiving and using information

[1238] The system receives the personal AI's response from the server and utilizes that information. For example, it receives specific advice such as, "Recommended fashion items for summer are a linen shirt and a denim skirt."

[1239] 3. Receiving emotional responses

[1240] When a user enters a query or feedback, the emotion engine recognizes the user's emotions and generates an appropriate response. For example, if the user expresses dissatisfaction by saying "this advice was not helpful," the engine will provide additional advice.

[1241] 4. Providing Feedback

[1242] Users can input their thoughts on the information provided or any follow-up questions as feedback and provide it to the server. The server then collects the feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows the influencer to more accurately understand the needs and sentiments of users and improve their content.

[1243] Examples of prompt statements

[1244] "Please tell me about your summer fashion coordination."

[1245] "This idea was completely useless."

[1246] "Please tell me some more casual outfit ideas."

[1247] This will streamline communication between influencers and users, and by taking emotions into consideration, more personalized information will be provided.

[1248] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1249] Step 1:

[1250] Content Collection

[1251] The server collects influencers' existing content through APIs and web scraping. Specifically, it uses the YouTube API to obtain video data and social media APIs to collect text and image data. The server specifies the influencer's account or specific keywords as input, and obtains video data, text data, and image data as output.

[1252] What it does: The server periodically sends a YouTube API request to get the metadata and video URL of new videos, then uses another API request to download the videos.

[1253] Step 2:

[1254] Data Preprocessing

[1255] The server preprocesses the acquired data. Specifically, it converts the video audio into text (e.g., using FFmpeg), extracts features from the image data (e.g., using TensorFlow), and cleanses the text data (e.g., removing unnecessary characters and tags using regular expressions). It receives video, text, and image data as input and generates preprocessed data as output.

[1256] What it does: It converts the video into an audio file using FFmpeg, then converts the audio to text using the Google Cloud Speech-to-Text API, and then formats the text using a natural language processing library.

[1257] Step 3:

[1258] Training generative AI models

[1259] The server uses the preprocessed data to train a generative AI model. Specifically, it uses GPT-3, for example, to learn style and expertise from influencer content. It receives the preprocessed data as input and generates a generative AI model as output.

[1260] Specific operation: The server uses TensorFlow or PyTorch to train a generative AI model and saves the model after learning is complete.

[1261] Step 4:

[1262] Receiving a user request

[1263] The server receives user inquiries via a REST API. For example, it receives a request such as, "Tell me about summer fashion coordination." It receives the user's question as input and sends the request to the appropriate personal AI as output.

[1264] What happens: The user's request is sent to the server as an HTTP POST request and parsed.

[1265] Step 5:

[1266] Generate and send a response

[1267] The server passes the received request to the generative AI model, which then generates an appropriate answer. For example, it might generate an answer such as "I recommend a linen shirt and a denim skirt," and the server sends that answer to the user's device. The generative AI model receives the user request and the generated answer as input and sends the generated answer to the user as output.

[1268] How it works: The generative AI model generates a response, and the server displays the response on the user's chat screen in real time.

[1269] Step 6:

[1270] Emotion analysis

[1271] The server uses an emotion engine to analyze the text of user inquiries and feedback and recognize the user's emotions. For example, if the user gives negative feedback such as "This idea was completely useless," it will identify it as "anger." It receives user feedback as input and generates emotion data as output.

[1272] Specific operation: Extracts emotions from text using a sentiment analysis algorithm and assigns sentiment tags.

[1273] Step 7:

[1274] Regulating responses based on emotions

[1275] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We're sorry for the inconvenience. What do you think of these suggestions?" and sends it to the user. It receives the sentiment data and the original answer as input and generates the adjusted response as output.

[1276] Specific operation: Based on the emotion data, a new prompt is given to the generative AI model to generate an appropriate response.

[1277] Step 8:

[1278] Collecting and summarizing feedback

[1279] The server collects user feedback and uses generative AI to summarize it. The feedback includes emotional information and provides it to influencers, who can then create new content based on user needs. The server receives user feedback as input and generates summarized feedback as output.

[1280] Specific operation: Collect multiple feedback data, generate a summary using natural language processing technology, and notify the influencer of the summary.

[1281] (Application example 2)

[1282] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1283] In recent years, as the influence of influencers has grown, there has been an increasing demand for information provision systems that utilize influencer content. However, existing systems have difficulty generating appropriate responses that take user emotions into account, limiting the improvement of user experience. There is also a growing need for systems that not only generate personal AI that reflects the style of a specific influencer, but also have the ability to adjust responses according to different emotional states. Furthermore, there is a need for a mechanism that allows users to receive more personalized information in virtual stores. The present invention aims to solve these problems and provide a high-quality information provision system that reflects user emotions.

[1284] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting existing content of influencers, means for training a generative AI model based on the collected existing content, means for generating a personal AI for each influencer using the generative AI model, means for receiving inquiries from users, means for selecting an optimal personal AI based on the inquiries, means for sending answers generated by the personal AI to the user, means for collecting feedback from users, means for summarizing the feedback and providing it to the influencer, means for analyzing the user's emotions, and means for adjusting the response of the personal AI based on the analyzed emotions. This makes it possible to adjust responses according to the user's emotional state, thereby enabling the provision of more personalized, high-quality information.

[1285] An "influencer" is someone who has a large following and whose opinions and actions influence other people.

[1286] "Existing content" refers to data such as text, images, and videos that influencers have previously posted.

[1287] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task based on collected data.

[1288] "Personal AI" refers to artificial intelligence generated to reflect the style and expertise of a specific influencer.

[1289] "Enquiry" means a question or request that a user enters into a system seeking information.

[1290] A "response" is an answer that a personal AI provides to a user's inquiry.

[1291] "Feedback" refers to the evaluations and opinions that users give in response to suggestions and advice.

[1292] "Sentiment analysis" is the process of analyzing and identifying a user's emotional state from their text and behavior.

[1293] "Response adjustment" is the process of appropriately modifying the answers provided by personal AI based on analyzed emotional data.

[1294] This invention is a system that collects existing influencer content, creates a generative AI model, and generates a personal AI for each influencer, which provides appropriate responses to user inquiries. It also analyzes user emotions and adjusts responses based on that information, thereby providing more personalized information.

[1295] Server-side implementation

[1296] The server first collects existing content from influencers through APIs and web scraping. The collected data is stored in a database, where it undergoes pre-processing before being fed into a generative AI model. This generative AI model learns the style and expertise of a specific influencer and generates a personal AI based on the results.

[1297] Inquiries from users are received via REST APIs, and the most suitable personal AI is selected based on the content of the inquiry. The selected personal AI generates a response to the inquiry and sends it from the server to the user's device. At this time, the server is equipped with an emotion analysis engine that analyzes the user's emotions and adjusts the response based on that information.

[1298] By utilizing the sentiment analysis engine, if a user sends negative feedback, such as "This advice was not helpful," the sentiment engine will detect "anger" and the personal AI will adjust its response to be more polite. This response adjustment process may include providing additional advice or citing past success stories.

[1299] Terminal side embodiment

[1300] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. Questions entered by the user are sent from the device to the server, and responses received from the server are immediately displayed on the chat screen.

[1301] The feedback sent from the device includes satisfaction ratings and emotional data. This information is sent to a server for analysis. The emotional data is fed back into the generative AI model, helping to improve the accuracy of each influencer's personal AI.

[1302] User's embodiment

[1303] Users operate the application and start a chat when they want information from a specific influencer. They receive a response to their input from a personal AI, which then utilizes that information. For example, if a user asks, "I'd like some advice on casual summer style," they can receive specific advice from the influencer AI, such as, "I recommend a linen shirt and a denim skirt."

[1304] The feedback provided by users includes information based on sentiment analysis. For example, if a user is confused and says, "I want more specific suggestions," the system will provide additional, detailed advice. This feedback is collected on the server and provided to influencers, allowing them to generate new content based on user needs.

[1305] Examples and prompts

[1306] As a concrete example, imagine a scenario in which a user is choosing fashion in a virtual store. For example, if the user asks, "I'd like some advice on a casual summer style," the system will respond, "I recommend a linen shirt and a denim skirt."

[1307] Example prompt sentence:

[1308] "A user is asking about casual summer style. What suggestions can you offer?"

[1309] "The user is confused. Please provide more specific advice."

[1310] This allows them to take into account the user's emotions and provide a more personalized response.

[1311] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1312] Step 1:

[1313] The server collects existing content from influencers through APIs and web scraping, and stores it in a database in the form of video, text, images, etc.

[1314] Input: Influencer content (e.g., video, text, images)

[1315] Output: Content data stored in the database

[1316] How it works: The server uses web scraping tools and APIs to retrieve content from social media and video platforms and stores it in a database.

[1317] Step 2:

[1318] The server preprocesses the collected data and feeds it into the generative AI model. Data preprocessing includes removing unnecessary information and standardizing the format.

[1319] Input: Content data in the database

[1320] Output: Preprocessed data

[1321] Specific operation: The server organizes the data using data cleaning techniques and performs preprocessing such as tokenizing text and resizing images.

[1322] Step 3:

[1323] The server uses the preprocessed data to train a generative AI model, which acts as a personal AI.

[1324] Input: Preprocessed data

[1325] Output: A trained generative AI model

[1326] How it works: The server runs machine learning algorithms and trains generative AI models based on the collected data.

[1327] Step 4:

[1328] The user's device starts the application and displays the login screen. The user logs in.

[1329] Input: User login information

[1330] Output: User authentication result

[1331] Specific operation: The terminal sends the login information entered by the user to the server, and the server performs authentication.

[1332] Step 5:

[1333] The terminal displays a chat screen and receives inquiries from users.

[1334] Input: User's inquiry

[1335] Output: Inquiry content displayed on the chat screen

[1336] Specific behavior: Sends the text entered by the user on the chat screen to the server.

[1337] Step 6:

[1338] The server receives the user's inquiry and selects the most suitable personal AI.

[1339] Input: User's inquiry

[1340] Output: Selected personal AI

[1341] Specific operation: The server analyzes the query content and selects the most suitable personal AI model.

[1342] Step 7:

[1343] The server sends the response generated by the selected personal AI to the user's device.

[1344] Input: Enquiry details, personal AI model

[1345] Output: Response by personal AI

[1346] Specific operation: The personal AI generates a response based on the inquiry, and the server sends that response to the user's device.

[1347] Step 8:

[1348] The server receives the user's feedback and analyzes its content and sentiment.

[1349] Input: User feedback

[1350] Output: Parsed emotion data

[1351] Specific operation: The emotion analysis engine built into the server analyzes the feedback text to determine the emotion.

[1352] Step 9:

[1353] The server adjusts the personal AI's responses based on the analyzed emotions.

[1354] Input: Emotional data, Personal AI response

[1355] Output: Adjusted response

[1356] Specific behavior: The server takes into account the emotional data and appropriately modifies the personal AI's responses, for example, making the response more polite if anger is detected.

[1357] Step 10:

[1358] The server summarizes the user's feedback and stores it for providing to the influencer.

[1359] Input: User feedback

[1360] Output: Summarized feedback data

[1361] Specific operation: The server uses generative AI to summarize the feedback content and saves the results in a database.

[1362] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1363] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1364] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1365] [Fourth embodiment]

[1366] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1367] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1368] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1369] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1370] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1371] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1372] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1373] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1374] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1375] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1376] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1377] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1378] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1379] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[1380] Server-side implementation

[1381] 1. Content collection and learning

[1382] The server collects influencers' existing content through APIs and web scraping and stores it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is preprocessed and fed into a generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[1383] 2. Providing chat-style responses

[1384] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[1385] 3. Gathering and Summarizing Feedback

[1386] The server collects feedback from users, including their question history, comments, and impressions. Generative AI is used to summarize the feedback and provide it to influencers in a concise format, allowing the influencers to create new content based on user needs.

[1387] Terminal side embodiment

[1388] 1. Use of the Application

[1389] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[1390] 2. Operations on the chat screen

[1391] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1392] 3. Providing Feedback

[1393] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[1394] User's embodiment

[1395] 1. Start a chat

[1396] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[1397] 2. Receiving and using information

[1398] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[1399] 3. Providing Feedback

[1400] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using generative AI, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[1401] This will streamline communication between influencers and users and promote the circulation of sustainable content.

[1402] The processing flow will be explained below.

[1403] Server-side processing

[1404] Step 1:

[1405] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[1406] Step 2:

[1407] The server pre-processes the collected content and feeds it into the generative AI model. Pre-processing includes text cleansing (noise removal), tokenization (breaking down words and sentences), and data batching.

[1408] Step 3:

[1409] The server trains a generative AI model to generate a personal AI that reflects the style and expertise of a specific influencer by running feedforward and backpropagation on the model using a large dataset.

[1410] Step 4:

[1411] The server receives inquiries from users via REST APIs, etc. The inquiry content is sent to the server in text format.

[1412] Step 5:

[1413] The server selects an appropriate personal AI based on the user's inquiry. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[1414] Step 6:

[1415] The selected personal AI generates a response to the query, and the generated text is sent to the server and prepared as the final answer.

[1416] Step 7:

[1417] The server then sends the generated answer to the user's device, where it is displayed in the user's application.

[1418] Step 8:

[1419] The server collects feedback from users, which is received in text format.

[1420] Step 9:

[1421] The server uses generative AI to summarize the collected feedback, which involves briefly summarizing the feedback and categorizing positive comments and requests for improvement.

[1422] Step 10:

[1423] The server provides summarized feedback to the influencer, who then uses it to plan and create new content.

[1424] Terminal side processing

[1425] Step 1:

[1426] A system-specific application is installed on the user's device.

[1427] Step 2:

[1428] The user launches the application and logs in by entering their account information.

[1429] Step 3:

[1430] Once the application has logged in, it will display the chat screen.

[1431] Step 4:

[1432] Questions and requests entered by the user are sent from the application to the server.

[1433] Step 5:

[1434] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[1435] Step 6:

[1436] If the user provides feedback, the application sends the feedback to the server.

[1437] User Action

[1438] Step 1:

[1439] The user launches the application and logs in.

[1440] Step 2:

[1441] Users go to the chat screen and type in their questions or requests.

[1442] Step 3:

[1443] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[1444] Step 4:

[1445] When users provide feedback, they type their thoughts about the answer to their question or any follow-up questions into the chat screen.

[1446] Step 5:

[1447] Send feedback and have it delivered to the server.

[1448] This will enable efficient communication between influencers and users, realizing a sustainable circulation of content.

[1449] Example 1

[1450] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1451] Currently, there is a lack of systems that can efficiently collect influencer content and provide personalized information to users. There is also a lack of efficient methods for conveying user feedback to influencers. This makes communication between users and influencers difficult, making it difficult to create new content based on user needs.

[1452] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1453] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiry; means for sending responses generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for storing the collected content in a database; means for using natural language processing technology to preprocess the collected data; means for inputting the preprocessed data as training data into the generative AI model; means for executing the generative AI model to generate responses; and means for summarizing the collected feedback using natural language processing technology. This not only effectively collects existing content from influencers, generates a personal AI based on the collected content, and provides personalized responses to users, but also efficiently communicates user feedback to influencers, facilitating the creation of new content.

[1454] An "influencer" is a person who has influence mainly on social media and online platforms and disseminates their opinions and information to a large number of followers.

[1455] "Content" is a general term for information, including data formats such as text, images, and videos, that are produced and disseminated by influencers.

[1456] A "generative AI model" is an artificial intelligence technology that uses machine learning algorithms to learn from large amounts of data and generate text that is as natural as a human's.

[1457] "Personal AI" is a generative AI model that learns the style and expertise of a specific influencer and generates information similar to that influencer.

[1458] A "query" is a question or request that a user enters into a system to ask for specific information.

[1459] "Feedback" refers to the user's reaction to the information provided, such as rating, comments, or additional questions.

[1460] A "database" is a system for efficiently storing, managing, and searching collected data.

[1461] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, including text cleaning, tokenization, and labeling.

[1462] "Training data" is a pre-processed data set used to train a generative AI model.

[1463] A "REST API" is a type of web service that is an interface for exchanging information between a server and a client using the HTTP protocol.

[1464] The present invention provides a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in a chat format. The following describes an embodiment of the system in detail.

[1465] Server-side implementation

[1466] The server uses APIs and web scraping techniques to collect existing content from influencers. Specifically, it uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The collected data is then stored in a NoSQL database such as MongoDB.

[1467] The server then preprocesses the collected data using natural language processing techniques, such as NLP libraries like SpaCy and NLTK. Preprocessing involves cleaning, tokenizing, and labeling the text. The video data is then converted to text using speech recognition techniques (e.g., Google Cloud Speech-to-Text API).

[1468] The preprocessed data is then used to train a generative AI model, such as GPT-3 or BERT, using machine learning frameworks like PyTorch or TensorFlow to generate a personalized AI model that reflects the style and expertise of a specific influencer.

[1469] The server receives user queries via a REST API. Frameworks such as Flask and FastAPI are used for this. If a user asks, "Tell me about summer fashion coordination," the request is sent to the personal AI, which generates a response. The response is sent back to the server, which then sends it to the user's device.

[1470] In addition, the server collects feedback from users, which is summarized using NLP technology and stored in a database. This summarized feedback is provided to influencers to help them create new content based on user needs.

[1471] Terminal side embodiment

[1472] A mobile application dedicated to this system is installed on the user's device. This application was developed using React Native and Flutter. The user launches the app and logs in. Once authentication is successful, a chat screen appears, allowing the user to freely enter questions or requests.

[1473] When a user types a question into the chat screen, the data is sent from the device to the server. For example, a user might type, "What are the latest summer fashion trends?" and press the send button. This request is sent to the server, which then forwards it to the appropriate personal AI. The generated response is sent from the server to the device and displayed on the chat screen. For example, specific advice such as, "I recommend a linen shirt and a denim skirt" is displayed.

[1474] Users can input their thoughts about the information provided and any follow-up questions as feedback, which is then sent from the device to the server. This feedback also includes a satisfaction rating. The device then sends this information in an appropriate format, making it easier for the server to analyze.

[1475] User's embodiment

[1476] Users interact with the app and initiate a chat to request information from a specific influencer or to find information related to a specific topic, for example, by typing in a question like "summer fashion" and waiting for a response to their question.

[1477] The system receives the answers sent by the influencer AI from the server and utilizes that information. For example, you can receive specific fashion advice such as "Linen shirts and denim skirts are trending."

[1478] Finally, users can provide feedback to the system, including their thoughts and ratings on the information provided. This feedback is then used by the system and influencers to create more optimized content.

[1479] Prompt Sentence Examples

[1480] As a concrete example, suppose a user enters the following question into an application:

[1481] "Please tell me about your summer fashion coordination."

[1482] An example prompt sentence to generate a response to this request:

[1483] "Provide summer fashion recommendations. Users want casual, comfortable outfits."

[1484] By inputting this prompt into a generative AI model, the influencer AI can provide specific suggestions to users, such as "I recommend a linen shirt, a denim skirt, and lightweight sandals."

[1485] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1486] System program processing flow and specific operations

[1487] Step 1:

[1488] The server collects existing content from influencers. It uses Python's BeautifulSoup library to extract text data from websites and YouTube API to collect video data. The input is the website URL and YouTube video ID, and the output is the extracted text data and downloaded video data. The collected data is stored in MongoDB.

[1489] Step 2:

[1490] The server preprocesses the collected data. It uses an NLP library (e.g., SpaCy or NLTK) to clean, tokenize, and label the text. It then uses the Google Cloud Speech-to-Text API to extract audio from the video data and convert it to text. The input is the raw data collected in step 1, and the output is the preprocessed text data.

[1491] Step 3:

[1492] The server inputs the preprocessed data into a generative AI model and trains the model. For example, it uses PyTorch or TensorFlow to train a model like GPT-3 or BERT. The input is preprocessed text data, and the output is a personal AI model that reflects the style and expertise of a specific influencer.

[1493] Step 4:

[1494] A user launches the application on their device and logs in. The app is developed using React Native and Flutter. If user authentication is successful, a chat screen is displayed. The input is the user's authentication information, and the output is the success / failure status of the authentication and the display of the chat screen.

[1495] Step 5:

[1496] The user enters a question or request on the chat screen, and the device sends the data to the server. For example, a question might be, "Tell me about summer fashion coordination." The input is the text of the question entered by the user, and the output is the request sent to the server.

[1497] Step 6:

[1498] The server sends the received user inquiry to the appropriate personal AI and generates a response. The generated response is received again by the server and sent to the user's device. For example, the personal AI generates an answer such as "I recommend a linen shirt and a denim skirt." The input is the text of the user's inquiry, and the output is the text of the response generated by the personal AI.

[1499] Step 7:

[1500] The terminal receives the response sent from the server and displays it on the chat screen. The user can refer to the displayed information. For example, the advice "We recommend a linen shirt and a denim skirt" is displayed. The input is the text of the response received from the server, and the output is to display that text on the chat screen.

[1501] Step 8:

[1502] The user inputs feedback on the provided information and sends it to the server from the terminal. The feedback includes ratings and follow-up questions. The input is the text of the feedback entered by the user, and the output is the feedback sent to the server.

[1503] Step 9:

[1504] The server summarizes the collected feedback using natural language processing technology and stores it in a database. The summarized feedback is provided to influencers. The input is the text of the feedback received from users, and the output is the summarized text of the feedback. This allows influencers to use it to create new content.

[1505] This will facilitate smooth communication between users and influencers, making it possible to create new content based on user needs.

[1506] (Application example 1)

[1507] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1508] In recent years, there has been an increasing demand for marketing and content provision using influencers. However, there are limited ways for users to receive fast, personalized information from specific influencers, making it difficult for influencers to accurately understand user needs and create new content. Furthermore, existing content provided in the form of videos, images, and text makes it difficult for users to obtain information in real time and lacks interactivity. A new system is needed to solve this issue and streamline communication between users and influencers.

[1509] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1510] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the user; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for conducting real-time dialogue via a smartphone or head-mounted display; and means for generating answers in response to questions from users and displaying them in AR or chat format. This allows users to receive information in real time from specific influencer AIs, enabling influencers to create new content based on user needs.

[1511] An "influencer" is an individual or group who influences a large number of followers through an online platform.

[1512] "Existing content" refers to digital information such as text, images, and videos that have already been created and published by influencers.

[1513] A "generative AI model" is an artificial intelligence that is trained using natural language processing and machine learning techniques based on collected data.

[1514] "Personal AI" is a generative AI model that is customized to reflect the style and expertise of a specific influencer.

[1515] An "inquiry" is a question or request for information that a user makes to the influencer AI.

[1516] A "smartphone" is a portable electronic device that, in addition to the functions of a mobile phone, has advanced computing power and Internet access capabilities.

[1517] A "head-mounted display" is a display device worn on the user's head, typically for displaying visual and auditory content.

[1518] "Real-time" refers to processing and responding to information and data instantly, without delay.

[1519] "Chat format" refers to a format in which a conversation is carried out between the user and the system by sending and receiving text messages.

[1520] "AR (Augmented Reality)" is a technology that displays digital information overlaid on images of the real world.

[1521] "Feedback" refers to reaction information based on comments, impressions, ratings, etc. provided by users.

[1522] The present invention is a system that collects existing content from influencers, trains a generative AI model to generate a personal AI, and provides information to users in chat or AR formats. The following describes an embodiment of the system in detail.

[1523] Server-side implementation

[1524] 1. Content collection and learning

[1525] The server uses APIs and web scraping technology to collect influencers' existing content. This content includes video data from video platforms and text and image data from social media. This data is stored in a database and, after preprocessing, is used as training data for the generative AI model. The generative AI model learns from this data and generates a personal AI that reflects the style and expertise of a specific influencer.

[1526] 2. Providing chat and AR responses

[1527] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer based on the inquiry, and the server sends the answer to the user's device. The answer is displayed on the user's device in chat format or AR format.

[1528] 3. Gathering and Summarizing Feedback

[1529] The server collects feedback from users, including their question history, comments, and impressions. It then uses a generative AI model to summarize the feedback and provide it to influencers, who can then create new content based on user needs.

[1530] Terminal side embodiment

[1531] 1. Use of the Application

[1532] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. If a head-mounted display is used, information is provided in AR format.

[1533] 2. Operations on the chat screen and AR screen

[1534] The questions entered by the user are sent from the device to the server. The responses received from the server are instantly displayed on the chat screen or AR screen, and the user can use the information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1535] 3. Providing Feedback

[1536] The feedback provided by the user is sent from the device to the server. This feedback can include new questions, comments, satisfaction ratings, etc. The device sends this information in an appropriate format to facilitate analysis on the server side.

[1537] User's embodiment

[1538] 1. Start a chat

[1539] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[1540] 2. Receiving and using information

[1541] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[1542] 3. Providing Feedback

[1543] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server collects this feedback, summarizes it using a generative AI model, and provides it to the influencer. This process allows influencers to more accurately understand users' needs and improve their content.

[1544] Specific examples

[1545] If the user asks "Tell me about summer fashion," the following prompt will be generated:

[1546] "Users ask AI fashion influencers questions: Tell us about summer fashion"

[1547] This allows users to receive specific fashion advice in chat or AR format.

[1548] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1549] Step 1:

[1550] The server collects influencers' existing content. Specifically, it uses APIs and web scraping technology to collect text, image, and video data from video platforms and social media, and stores it in a database. The collected data undergoes preprocessing (noise removal, normalization, tokenization, etc.) and is prepared as training data for the generative AI model. The input is public content on the internet, and the output is preprocessed learning data.

[1551] Step 2:

[1552] The server uses the preprocessed data to train a generative AI model using natural language processing or machine learning algorithms (e.g., the GPT series). The input is the preprocessed content data, and the output is a generative AI model that has learned the style and expertise of a specific influencer. Specifically, the server splits the dataset and iteratively trains the model.

[1553] Step 3:

[1554] The server receives inquiries from users via a REST API. For example, if a user asks, "Tell me about summer fashion coordination," the server converts the inquiry into an appropriate format. The input is the user's inquiry text, and the output is data that has been appropriately analyzed from the inquiry. Specifically, the server uses natural language processing to tokenize and analyze the inquiry.

[1555] Step 4:

[1556] The server selects the optimal personal AI based on the content of the inquiry. This generates a response based on the style and expertise of a specific influencer. The input is the analyzed inquiry content and information from multiple personal AIs, and the output is the optimal personal AI. Specifically, it compares the characteristics of the inquiry content and the personal AIs and selects the AI ​​with the highest degree of match.

[1557] Step 5:

[1558] The server generates a response to the inquiry using the selected personal AI. The generated answer is sent to the user. The input is the inquiry content and the selected personal AI, and the output is the generated answer. Specifically, a prompt sentence is input into the generative AI model to obtain a natural conversational response.

[1559] Step 6:

[1560] The device displays the answers received from the server on the chat screen or in AR format. The input is the answer data sent from the server, and the output is the information visually displayed to the user. Specifically, the information is displayed in text format or augmented reality format on the screen of the chat application or AR display.

[1561] Step 7:

[1562] The user inputs feedback on the provided information. The input is the user's feedback text, and the output is the feedback data sent to the server. Specifically, the information is entered and sent from a feedback input form in the chat application.

[1563] Step 8:

[1564] The server collects feedback from users and summarizes it using a generative AI model. The input is the feedback data sent by the user, and the output is the summarized feedback information. Specifically, the server analyzes the feedback data using natural language processing technology, extracts key points, and generates a summary.

[1565] Step 9:

[1566] The server provides summarized feedback to influencers. The input is the summarized feedback data, and the output is the data provided in the form of notifications or reports to the influencers. Specific actions include sending the feedback to the influencers' dashboards or notification systems.

[1567] The above are the specific processing steps and their contents of the system that realizes the application example.

[1568] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1569] The present invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, provides information to users in a chat format, and combines it with an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions. The following describes an embodiment of the system in detail.

[1570] Server-side implementation

[1571] 1. Content collection and learning

[1572] The server first collects the influencer's existing content through APIs and web scraping, storing it in a database. For example, it collects video data from video platforms and text and image data from social media. This data is then preprocessed and fed into a generative AI model. The generative AI model then learns from this data and generates a personal AI that reflects the style and expertise of the specific influencer.

[1573] 2. Providing chat-style responses

[1574] The server receives inquiries from users via REST APIs, etc. For example, if a user asks, "Tell me about summer fashion coordination," the server sends the request to the appropriate personal AI. The personal AI generates an answer in response to the inquiry, and the server sends the answer to the user's device.

[1575] 3. Emotion engine integration

[1576] The server is equipped with an emotion engine that analyzes user inquiries and feedback to recognize user emotions, such as joy, anger, sadness, and surprise, based on the context of the text.

[1577] 4. Emotion-Based Response Modulation

[1578] The server adjusts the personal AI's response based on the user's emotional information analyzed by the emotion engine. For example, if a user gives negative feedback such as "This idea was completely useless," the emotion engine will determine this as "anger," and the personal AI will adjust its response to be more polite, such as providing additional advice or citing past success stories.

[1579] 5. Collecting and Summarizing Feedback

[1580] The server collects user feedback and uses generative AI to summarize it. The summary, including the sentiment data of the feedback, is provided to influencers in a concise format, allowing them to create new content based on user needs.

[1581] Terminal side embodiment

[1582] 1. Use of the Application

[1583] A dedicated application for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests.

[1584] 2. Operations on the chat screen

[1585] The questions entered by the user are sent from the device to the server. The response received from the server is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1586] 3. Providing Feedback

[1587] The feedback provided by the user is transmitted from the device to the server. This feedback includes new questions, comments, satisfaction ratings, and even emotional data. The device transmits this information in an appropriate format to facilitate analysis on the server side.

[1588] User's embodiment

[1589] 1. Start a chat

[1590] Users interact with the application to initiate a chat to request information from a specific influencer or to find information related to a specific topic. They type in a question and wait for a response to that question.

[1591] 2. Receiving and using information

[1592] The system receives the answers of the influencer AI sent from the server and utilizes that information. For example, if a user asks "summer fashion," they can receive appropriate fashion advice from the influencer AI.

[1593] 3. Receiving emotional responses

[1594] Based on the user's input query and feedback, the system uses an emotion engine to generate an appropriate response. For example, if the user is confused, the system will provide additional explanation.

[1595] 4. Providing Feedback

[1596] Users can provide feedback, such as their thoughts on the information provided or any follow-up questions. The server then collects this feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows influencers to more accurately understand users' needs and emotions and improve their content.

[1597] This will streamline communication between influencers and users, and promote the circulation of more personalized and sustainable content by taking emotions into consideration.

[1598] The processing flow will be explained below.

[1599] Server-side processing

[1600] Step 1:

[1601] The server collects influencers' existing content through APIs and web scraping and stores it in a database, for example, by collecting video data from video platforms and text and image data from social media.

[1602] Step 2:

[1603] The server pre-processes the collected content, specifically cleansing, tokenizing, and batching the text data to prepare it for ingestion into the generative AI model.

[1604] Step 3:

[1605] The server trains a generative AI model, using the pre-processed content as training data and running feedforward and backpropagation on the model to generate a personal AI that reflects the style and expertise of a particular influencer.

[1606] Step 4:

[1607] The server receives inquiries from users via the REST API. For example, the user might ask, "Tell me about summer fashion coordination."

[1608] Step 5:

[1609] The server uses an emotion engine to analyze the content of the user's inquiry and recognize their emotion. For example, it can determine their interest from the context of "Please tell me."

[1610] Step 6:

[1611] The server selects an appropriate personal AI based on the user's inquiry and emotional information. If a specific influencer is specified, that influencer AI will be selected. If no influencer is specified, the server will use natural language processing (NLP) technology to select the most suitable influencer AI.

[1612] Step 7:

[1613] The selected personal AI generates an answer based on the user's query and emotional information, and the generated text is sent to the server to be prepared as the final answer.

[1614] Step 8:

[1615] The server then sends the generated answer to the user's device, such as "A cool linen shirt and denim skirt are recommended for this summer."

[1616] Step 9:

[1617] The server collects feedback from users. The feedback is received in text format, and the emotion engine simultaneously analyzes the emotion data. For example, feedback such as "This was helpful" is recognized as "satisfied / positive."

[1618] Step 10:

[1619] The server summarizes the collected feedback and its sentiment data, then uses generative AI to summarize it and provide it to influencers in a concise format, allowing them to create new content that reflects user needs and sentiment.

[1620] Terminal side processing

[1621] Step 1:

[1622] The terminal installs a system-specific application.

[1623] Step 2:

[1624] The user launches the application and logs in by entering their account information.

[1625] Step 3:

[1626] Once the application has logged in, it will display the chat screen.

[1627] Step 4:

[1628] Questions and requests entered by the user are sent from the device to the server.

[1629] Step 5:

[1630] Once the response is received from the server, the application displays it on the chat screen, where the user can view the information.

[1631] Step 6:

[1632] When the user provides feedback, the application sends the feedback to the server, along with the user's emotional data.

[1633] User Action

[1634] Step 1:

[1635] The user launches the application and logs in.

[1636] Step 2:

[1637] Users go to the chat screen and type in their questions or requests.

[1638] Step 3:

[1639] The user checks the response from the server on the chat screen, and once the response is displayed, the user can use the information.

[1640] Step 4:

[1641] When users provide feedback, they type their thoughts about the answer or follow-up questions into the chat window, and emotional data is automatically analyzed and sent along with the feedback.

[1642] Step 5:

[1643] Send feedback and have it delivered to the server, where it will be analyzed and summarized by the system and provided to the influencer.

[1644] This will enable more efficient communication between influencers and users, and by taking emotions into consideration, will enable a more personalized and sustainable circulation of content.

[1645] Example 2

[1646] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1647] In the conventional system, information provided through influencer content was one-way, making it difficult to respond flexibly to user emotions. Furthermore, there was no system in place to efficiently collect and analyze user feedback, making it difficult for influencers to accurately understand users' needs and emotions. This limited the user experience and prevented effective communication between influencers and users.

[1648] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1649] In this invention, the server includes: means for collecting existing content from influencers; means for training a generative AI model based on the collected existing content; means for generating a personal AI for each influencer using the generative AI model; means for receiving inquiries from users; means for selecting an optimal personal AI based on the inquiries; means for sending answers generated by the personal AI to the users; means for collecting feedback from users; means for summarizing the feedback and providing it to the influencer; means for preprocessing text data; means for analyzing emotions that recognize the user's emotions; and means for adjusting response content based on the recognized emotions. This enables flexible information provision based on user emotions and efficient collection and analysis of feedback.

[1650] An "influencer" is an individual or organization that uses social media or other digital platforms to influence a large number of followers.

[1651] "Existing content" refers to digital media data such as videos, text, and images that have already been published by influencers.

[1652] "Collection Methods" refers to the technical means used to obtain existing content from influencers, using technologies such as APIs and web scraping.

[1653] A "generative AI model" refers to an artificial intelligence model that learns from large amounts of data and generates natural language and images.

[1654] "Training method" refers to the technical process or method for training a generative AI model based on existing content.

[1655] "Personal AI" refers to artificial intelligence that learns the style and expertise of a specific influencer and provides information tailored to that influencer.

[1656] "Means for receiving inquiries" refers to the technical means by which the server receives inquiries from users, such as through a REST API.

[1657] "Means for selecting the optimal personal AI" refers to algorithms and technologies for selecting the appropriate personal AI based on the content of the user's inquiry.

[1658] "Means for sending answers" refers to the communication means for sending answers generated by personal AI to the user's device.

[1659] "Feedback Collection Means" refers to the technical means by which the Server collects comments and ratings provided by Users.

[1660] "Means for summarizing and providing feedback" refers to technology for summarizing collected feedback using a generative AI model and providing that summary to influencers.

[1661] "Means for preprocessing text data" refers to the technical means for analyzing and cleansing collected text data and converting it into a form that is easy for generative AI models to use.

[1662] "Sentiment analysis methods" refer to algorithms and technologies used to recognize and classify emotions from user text data.

[1663] "Means for adjusting response content" refers to technical means for appropriately changing the responses of personal AI based on the results of emotion analysis.

[1664] This invention relates to a system that generates a personal AI from a generative AI model that has been trained by collecting existing influencer content, and provides information to users in chat format, as well as a system that combines an emotion engine that recognizes the user's emotions and adjusts responses based on them.

[1665] Server-side implementation

[1666] 1. Content collection and learning

[1667] The server first collects the influencer's existing content through APIs and web scraping. For example, it obtains video data from video platforms and collects text and image data from social networking services. This data is preprocessed and fed into the generative AI model. Preprocessing includes converting the video's audio to text, extracting features from the image data, and cleansing the text. The generative AI model then learns from this data to generate a personal AI that reflects the style and expertise of a specific influencer. For example, it obtains video data using YouTube's API, extracts audio from the video using FFmpeg, and converts it into text using the Google Cloud Speech-to-Text API. The collected text data is then formatted using natural language processing technology.

[1668] 2. Receiving and Processing User Requests

[1669] When a user makes a request through the application, such as "Tell me about summer fashion coordination," the request is sent to the server via a REST API. The server selects an appropriate personal AI and communicates the inquiry to that personal AI. The generative AI model generates an answer in response to the inquiry. If the answer is "I recommend a linen shirt and a denim skirt," the server sends this answer to the user's device.

[1670] 3. Emotion engine integration and emotion analysis

[1671] The server is integrated with an emotion engine that analyzes user inquiries and feedback to recognize emotions. For example, if a user sends feedback such as "This idea was completely useless," the emotion engine analyzes the text and determines that the emotion is "anger." This is done using natural language processing technology and emotion classification algorithms.

[1672] 4. Adjusting the response

[1673] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We apologize for the inconvenience. What do you think of these suggestions?" and sends it to the user.

[1674] 5. Collecting and Summarizing Feedback

[1675] The server collects user feedback and uses generative AI to summarize it, including emotional information, and provides it to influencers, who can then create new content based on user needs.

[1676] Terminal side embodiment

[1677] 1. Use of the Application

[1678] Users install the dedicated application and log in. After logging in, a chat screen appears, allowing users to freely post questions or requests.

[1679] 2. Operations on the chat screen

[1680] When a user enters a question, the content is sent from the device to the server. The server's response is immediately displayed on the chat screen, and the user can use that information. For example, they can receive specific outfit suggestions such as, "We recommend a linen shirt and a denim skirt."

[1681] 3. Providing Feedback

[1682] Users can enter feedback on the information provided and send it from their device to the server. This feedback can include new questions, comments, satisfaction ratings, and sentiment data. The device then sends this information in an appropriate format to facilitate analysis by the server.

[1683] User's embodiment

[1684] 1. Start a chat

[1685] Users can navigate the app to start a chat to request information from a specific influencer or find information related to a specific topic, for example, by typing, "Tell me about summer fashion."

[1686] 2. Receiving and using information

[1687] The system receives the personal AI's response from the server and utilizes that information. For example, it receives specific advice such as, "Recommended fashion items for summer are a linen shirt and a denim skirt."

[1688] 3. Receiving emotional responses

[1689] When a user enters a query or feedback, the emotion engine recognizes the user's emotions and generates an appropriate response. For example, if the user expresses dissatisfaction by saying "this advice was not helpful," the engine will provide additional advice.

[1690] 4. Providing Feedback

[1691] Users can input their thoughts on the information provided or any follow-up questions as feedback and provide it to the server. The server then collects the feedback, summarizes it using generative AI, and collects sentiment data, which it then provides to the influencer. This process allows the influencer to more accurately understand the needs and sentiments of users and improve their content.

[1692] Examples of prompt statements

[1693] "Please tell me about your summer fashion coordination."

[1694] "This idea was completely useless."

[1695] "Please tell me some more casual outfit ideas."

[1696] This will streamline communication between influencers and users, and by taking emotions into consideration, more personalized information will be provided.

[1697] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1698] Step 1:

[1699] Content Collection

[1700] The server collects influencers' existing content through APIs and web scraping. Specifically, it uses the YouTube API to obtain video data and social media APIs to collect text and image data. The server specifies the influencer's account or specific keywords as input, and obtains video data, text data, and image data as output.

[1701] What it does: The server periodically sends a YouTube API request to get the metadata and video URL of new videos, then uses another API request to download the videos.

[1702] Step 2:

[1703] Data Preprocessing

[1704] The server preprocesses the acquired data. Specifically, it converts the video audio into text (e.g., using FFmpeg), extracts features from the image data (e.g., using TensorFlow), and cleanses the text data (e.g., removing unnecessary characters and tags using regular expressions). It receives video, text, and image data as input and generates preprocessed data as output.

[1705] What it does: It converts the video into an audio file using FFmpeg, then converts the audio to text using the Google Cloud Speech-to-Text API, and then formats the text using a natural language processing library.

[1706] Step 3:

[1707] Training generative AI models

[1708] The server uses the preprocessed data to train a generative AI model. Specifically, it uses GPT-3, for example, to learn style and expertise from influencer content. It receives the preprocessed data as input and generates a generative AI model as output.

[1709] Specific operation: The server uses TensorFlow or PyTorch to train a generative AI model and saves the model after learning is complete.

[1710] Step 4:

[1711] Receiving a user request

[1712] The server receives user inquiries via a REST API. For example, it receives a request such as, "Tell me about summer fashion coordination." It receives the user's question as input and sends the request to the appropriate personal AI as output.

[1713] What happens: The user's request is sent to the server as an HTTP POST request and parsed.

[1714] Step 5:

[1715] Generate and send a response

[1716] The server passes the received request to the generative AI model, which then generates an appropriate answer. For example, it might generate an answer such as "I recommend a linen shirt and a denim skirt," and the server sends that answer to the user's device. The generative AI model receives the user request and the generated answer as input and sends the generated answer to the user as output.

[1717] How it works: The generative AI model generates a response, and the server displays the response on the user's chat screen in real time.

[1718] Step 6:

[1719] Emotion analysis

[1720] The server uses an emotion engine to analyze the text of user inquiries and feedback and recognize the user's emotions. For example, if the user gives negative feedback such as "This idea was completely useless," it will identify it as "anger." It receives user feedback as input and generates emotion data as output.

[1721] Specific operation: Extracts emotions from text using a sentiment analysis algorithm and assigns sentiment tags.

[1722] Step 7:

[1723] Regulating responses based on emotions

[1724] The server adjusts the response content of the generative AI model based on the results of the sentiment analysis. For example, it generates a response such as "We're sorry for the inconvenience. What do you think of these suggestions?" and sends it to the user. It receives the sentiment data and the original answer as input and generates the adjusted response as output.

[1725] Specific operation: Based on the emotion data, a new prompt is given to the generative AI model to generate an appropriate response.

[1726] Step 8:

[1727] Collecting and summarizing feedback

[1728] The server collects user feedback and uses generative AI to summarize it. The feedback includes emotional information and provides it to influencers, who can then create new content based on user needs. The server receives user feedback as input and generates summarized feedback as output.

[1729] Specific operation: Collect multiple feedback data, generate a summary using natural language processing technology, and notify the influencer of the summary.

[1730] (Application example 2)

[1731] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1732] In recent years, as the influence of influencers has grown, there has been an increasing demand for information provision systems that utilize influencer content. However, existing systems have difficulty generating appropriate responses that take user emotions into account, limiting the improvement of user experience. There is also a growing need for systems that not only generate personal AI that reflects the style of a specific influencer, but also have the ability to adjust responses according to different emotional states. Furthermore, there is a need for a mechanism that allows users to receive more personalized information in virtual stores. The present invention aims to solve these problems and provide a high-quality information provision system that reflects user emotions.

[1733] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting existing content of influencers, means for training a generative AI model based on the collected existing content, means for generating a personal AI for each influencer using the generative AI model, means for receiving inquiries from users, means for selecting an optimal personal AI based on the inquiries, means for sending answers generated by the personal AI to the user, means for collecting feedback from users, means for summarizing the feedback and providing it to the influencer, means for analyzing the user's emotions, and means for adjusting the response of the personal AI based on the analyzed emotions. This makes it possible to adjust responses according to the user's emotional state, thereby enabling the provision of more personalized, high-quality information.

[1734] An "influencer" is someone who has a large following and whose opinions and actions influence other people.

[1735] "Existing content" refers to data such as text, images, and videos that influencers have previously posted.

[1736] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task based on collected data.

[1737] "Personal AI" refers to artificial intelligence generated to reflect the style and expertise of a specific influencer.

[1738] "Enquiry" means a question or request that a user enters into a system seeking information.

[1739] A "response" is an answer that a personal AI provides to a user's inquiry.

[1740] "Feedback" refers to the evaluations and opinions that users give in response to suggestions and advice.

[1741] "Sentiment analysis" is the process of analyzing and identifying a user's emotional state from their text and behavior.

[1742] "Response adjustment" is the process of appropriately modifying the answers provided by personal AI based on analyzed emotional data.

[1743] This invention is a system that collects existing influencer content, creates a generative AI model, and generates a personal AI for each influencer, which provides appropriate responses to user inquiries. It also analyzes user emotions and adjusts responses based on that information, thereby providing more personalized information.

[1744] Server-side implementation

[1745] The server first collects existing content from influencers through APIs and web scraping. The collected data is stored in a database, where it undergoes pre-processing before being fed into a generative AI model. This generative AI model learns the style and expertise of a specific influencer and generates a personal AI based on the results.

[1746] Inquiries from users are received via REST APIs, and the most suitable personal AI is selected based on the content of the inquiry. The selected personal AI generates a response to the inquiry and sends it from the server to the user's device. At this time, the server is equipped with an emotion analysis engine that analyzes the user's emotions and adjusts the response based on that information.

[1747] By utilizing the sentiment analysis engine, if a user sends negative feedback, such as "This advice was not helpful," the sentiment engine will detect "anger" and the personal AI will adjust its response to be more polite. This response adjustment process may include providing additional advice or citing past success stories.

[1748] Terminal side embodiment

[1749] An application specifically for this system is installed on the user's device. The user launches the application and logs in. After logging in, a chat screen appears, allowing the user to freely enter questions or requests. Questions entered by the user are sent from the device to the server, and responses received from the server are immediately displayed on the chat screen.

[1750] The feedback sent from the device includes satisfaction ratings and emotional data. This information is sent to a server for analysis. The emotional data is fed back into the generative AI model, helping to improve the accuracy of each influencer's personal AI.

[1751] User's embodiment

[1752] Users operate the application and start a chat when they want information from a specific influencer. They receive a response to their input from a personal AI, which then utilizes that information. For example, if a user asks, "I'd like some advice on casual summer style," they can receive specific advice from the influencer AI, such as, "I recommend a linen shirt and a denim skirt."

[1753] The feedback provided by users includes information based on sentiment analysis. For example, if a user is confused and says, "I want more specific suggestions," the system will provide additional, detailed advice. This feedback is collected on the server and provided to influencers, allowing them to generate new content based on user needs.

[1754] Examples and prompts

[1755] As a concrete example, imagine a scenario in which a user is choosing fashion in a virtual store. For example, if the user asks, "I'd like some advice on a casual summer style," the system will respond, "I recommend a linen shirt and a denim skirt."

[1756] Example prompt sentence:

[1757] "A user is asking about casual summer style. What suggestions can you offer?"

[1758] "The user is confused. Please provide more specific advice."

[1759] This allows them to take into account the user's emotions and provide a more personalized response.

[1760] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1761] Step 1:

[1762] The server collects existing content from influencers through APIs and web scraping, and stores it in a database in the form of video, text, images, etc.

[1763] Input: Influencer content (e.g., video, text, images)

[1764] Output: Content data stored in the database

[1765] How it works: The server uses web scraping tools and APIs to retrieve content from social media and video platforms and stores it in a database.

[1766] Step 2:

[1767] The server preprocesses the collected data and feeds it into the generative AI model. Data preprocessing includes removing unnecessary information and standardizing the format.

[1768] Input: Content data in the database

[1769] Output: Preprocessed data

[1770] Specific operation: The server organizes the data using data cleaning techniques and performs preprocessing such as tokenizing text and resizing images.

[1771] Step 3:

[1772] The server uses the preprocessed data to train a generative AI model, which acts as a personal AI.

[1773] Input: Preprocessed data

[1774] Output: A trained generative AI model

[1775] How it works: The server runs machine learning algorithms and trains generative AI models based on the collected data.

[1776] Step 4:

[1777] The user's device starts the application and displays the login screen. The user logs in.

[1778] Input: User login information

[1779] Output: User authentication result

[1780] Specific operation: The terminal sends the login information entered by the user to the server, and the server performs authentication.

[1781] Step 5:

[1782] The terminal displays a chat screen and receives inquiries from users.

[1783] Input: User's inquiry

[1784] Output: Inquiry content displayed on the chat screen

[1785] Specific behavior: Sends the text entered by the user on the chat screen to the server.

[1786] Step 6:

[1787] The server receives the user's inquiry and selects the most suitable personal AI.

[1788] Input: User's inquiry

[1789] Output: Selected personal AI

[1790] Specific operation: The server analyzes the query content and selects the most suitable personal AI model.

[1791] Step 7:

[1792] The server sends the response generated by the selected personal AI to the user's device.

[1793] Input: Enquiry details, personal AI model

[1794] Output: Response by personal AI

[1795] Specific operation: The personal AI generates a response based on the inquiry, and the server sends that response to the user's device.

[1796] Step 8:

[1797] The server receives the user's feedback and analyzes its content and sentiment.

[1798] Input: User feedback

[1799] Output: Parsed emotion data

[1800] Specific operation: The emotion analysis engine built into the server analyzes the feedback text to determine the emotion.

[1801] Step 9:

[1802] The server adjusts the personal AI's responses based on the analyzed emotions.

[1803] Input: Emotional data, Personal AI response

[1804] Output: Adjusted response

[1805] Specific behavior: The server takes into account the emotional data and appropriately modifies the personal AI's responses, for example, making the response more polite if anger is detected.

[1806] Step 10:

[1807] The server summarizes the user's feedback and stores it for providing to the influencer.

[1808] Input: User feedback

[1809] Output: Summarized feedback data

[1810] Specific operation: The server uses generative AI to summarize the feedback content and saves the results in a database.

[1811] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1812] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1813] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1814] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1815] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1816] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1817] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1818] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1819] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1820] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1821] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1822] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1823] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1824] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1825] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1826] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1827] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1828] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1829] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1830] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1831] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1832] The following is further disclosed regarding the above embodiment.

[1833] (Claim 1)

[1834] A means of collecting existing content from influencers and

[1835] means for training a generative AI model based on the collected existing content;

[1836] A means for generating a personal AI for each influencer using the generative AI model;

[1837] a means for receiving inquiries from users;

[1838] A means for selecting an optimal personal AI based on the inquiry;

[1839] means fo...

Claims

1. A means of collecting existing content from influencers and means for training a generative AI model based on the collected existing content; A means for generating a personal AI for each influencer using the generative AI model; a means for receiving inquiries from users; A means for selecting an optimal personal AI based on the inquiry; means for transmitting the answer generated by the personal AI to the user; a means of gathering user feedback; means for summarizing and providing said feedback to influencers; A system including:

2. The system of claim 1 , further comprising means for selecting a corresponding personal AI based on a user's designation of a particular influencer.

3. The system of claim 1 , further comprising means for providing optimal information across the personal AIs of multiple influencers.

Images